Reference compounds for alternative test methods to indicate developmental neurotoxicity (DNT) potential of chemicals: example lists and criteria for their selection and use.

HHS Public Access Author manuscript Author Manuscript

ALTEX. Author manuscript; available in PMC 2017 April 01. Published in final edited form as: ALTEX. 2017 ; 34(1): 49–74. doi:10.14573/altex.1604201.

Reference compounds for alternative test methods to indicate developmental neurotoxicity (DNT) potential of chemicals: example lists and criteria for their selection and use

Author Manuscript

Michael Aschner1, Sandra Ceccatelli2, Mardas Daneshian3, Ellen Fritsche4, Nina Hasiwa3, Thomas Hartung3,5, Helena T. Hogberg5, Marcel Leist3,6,7, Abby Li8, William R. Mundy9, Stephanie Padilla9, Aldert H. Piersma10,11, Anna Bal-Price12, Andrea Seiler13, Remco H. Westerink14, Bastian Zimmer15, and Pamela J. Lein16,17 1Albert

Author Manuscript

Einstein College of Medicine, New York, USA 2Department of Neuroscience, Karolinska Institutet Stockholm, Sweden 3Center for Alternatives to Animal Testing-Europe (CAAT-Europe), University of Konstanz, Germany 4Leibniz Research Institute for Environmental Medicine (IUF), Düsseldorf, Germany 5Center for Alternatives to Animal Testing (CAAT), The Johns Hopkins University, Baltimore, MD, USA 6Doerenkamp-Zbinden Chair for in vitro Toxicology and Biomedicine, University of Konstanz, Germany 7Konstanz Research School Chemical Biology (KoRS-CB), Konstanz University 8Exponent Inc., San Francisco, USA 9United States Environmental Protection Agency (US EPA), NHEERL, Research Triangle Park, NC, USA 10National Institute for Public Health and the Environment (RIVM), Bilthoven, The Netherlands 11Institute for Risk Assessment Sciences, Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands 12European Commission Joint Research Centre, Institute for Health and Consumer Protection, Ispra, Italy 13Federal Institute for Risk Assessment (BfR), Berlin, Germany 14Neurotoxicology Research Group, Institute for Risk Assessment Sciences (IRAS), Utrecht University, The Netherlands 15Axiogenesis AG, Köln, Germany 16Center for Research on Occupational and Environmental Toxicology, Oregon Health & Science University, Portland, USA 17Department of Molecular Biosciences, University of California, Davis, USA

Summary

Author Manuscript

There is a paucity of information concerning the developmental neurotoxicity (DNT) hazard posed by industrial and environmental chemicals. New testing approaches will most likely be based on batteries of alternative and complementary (non-animal) tests. As DNT is assumed to result from the modulation of fundamental neurodevelopmental processes (such as neuronal differentiation,

This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International license (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution and reproduction in any medium, provided the original work is appropriately cited. Correspondence to: Marcel leist, PhD, University of Konstanz, Box M657, 78457 Konstanz, Germany, [email protected]. Disclaimer: This article has been reviewed in accordance with the policy of the National Health and Environmental Effects Research Laboratory, U.S. Environmental Protection Agency, and approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use. Conflict of interest The authors declare that they have no conflicts of interest.

Aschner et al.

Page 2

Author Manuscript Author Manuscript

precursor cell migration or neuronal network formation) by chemicals, the first generation of alternative DNT tests target these processes. The advantage of such types of assays is that they capture toxicants with multiple targets and modes-of-action. Moreover, the processes modelled by the assays can be linked to toxicity endophenotypes, i.e. alterations in neural connectivity that form the basis for neurofunctional deficits in man. The authors of this review convened in a workshop to define criteria for the selection of positive/negative controls, to prepare recommendations on their use, and to initiate the setup of a directory of reference chemicals. For initial technical optimization of tests, a set of >50 endpoint-specific control compounds was identified. For further test development, an additional “test” set of 33 chemicals considered to act directly as bona fide DNT toxicants is proposed, and each chemical is annotated to the extent it fulfills these criteria. A tabular compilation of the original literature used to select the test set chemicals provides information on statistical procedures, and toxic/non-toxic doses (both for pups and dams). Suggestions are provided on how to use the >100 compounds (including negative controls) compiled here to address specificity, adversity and use of alternative test systems.

Keywords neurotoxicity; specificity; test development; AOP; validation

1 Introduction 1.1 DNT testing and test compound selection

Author Manuscript

Developmental neurotoxicity (DNT) may be broadly defined as an adverse change in the structure or function of the nervous system that manifests after exposure to a chemical during the prenatal or gestational period (Mundy et al., 2015). Notably, the adverse change canmanifest well after the toxicant exposure has ended, a phenomenon referred to as ‘delayed consequence of early life exposure’. This definition raises questions as to the type and magnitude of change considered to a relevant adverse effect. For practical purposes, any statistically significant change may be regarded as an alert for a potential DNT hazard, and then be followed up by more detailed studies. Most considerations of DNT focus on the ‘central nervous system’ but it may be questioned whether the peripheral nervous system, the gastrointestinal nervous system and/or other neural crest-derived tissues should be included in DNT studies.

Author Manuscript

Traditional approaches for generating data relevant to DNT hazard are largely based on animal testing, according to OECD TG426 and similar standardized protocols developed by national regulatory authorities. Such testing is time- and resource-consuming, which explains why currently only about 200 such studies have been performed with most directed towards pesticides and only a handful focused on industrial chemicals. Even amongst high production volume compounds, only a few have been studied for DNT hazards (Crofton et al., 2012; Rovida et al., 2011). It is also not clear whether these animal testing procedures are sufficiently sensitive to identify all hazardous substances that may affect the developing human brain. For instance, a guideline study on methylmercury, one of the best characterized DNT compounds that targets animals and man, failed to show adverse effects in rats when classical endpoints were condsidered. Only when specific imaging and

ALTEX. Author manuscript; available in PMC 2017 April 01.

Aschner et al.

Page 3

Author Manuscript

transcriptomics endpoints were included did this toxicant demonstrate adverse effects on the developing rat nervous system (Radonjic et al., 2013).

Author Manuscript

Epidemiological studies are an alternate approach for identifying DNT toxicants relevant to man. However, these studies can be particularly challenging due to the time lag between exposure and outcome measurement, and due to the multitude of potentially confounding factors (genetic variability, complex exposures, lifestyle factors, etc.) that affect the complex endpoints studied (e.g. neuropsychological, behavioural or cognitive performance tests). Until 2006, only 6 compounds (lead, mercury, arsenic, PCBs, toluene, ethanol) have been identified unambiguously by epidemiological approaches (Grandjean and Landrigan, 2006); further studies since then have expanded this list to include fluoride, manganese, tetrachloroethylene, chlorpyrifos, DDT, and PBDEs (Grandjean and Landrigan, 2006, 2014). Valproic acid needs to be added to this list based on clinical evidence (Kadereit, 2012; Balmer 2012). Thus, the total number of chemicals (n = 13) identified via clinical/ epidemiological studies is rather low to use as a reference chemical set for evaluating or establishing new test systems. Moreover, the epidemiological approach for identifying DNT chemicals provides negligible information as to whether these neurotoxic compounds are direct acting DNT compounds, and which neurodevelopmental processes are perturbed. Such knowledge is critically important for understanding how to use DNT test compounds for the evaluation and optimization of novel test systems. For example, toxicants acting on the thyroid may trigger DNT by decreasing thyroid hormone levels important for nervous system development, but such indirect effects would not be easily detectable in in vitro systems based on replicating specific neurodevelopmental processes.


Literature searches recently identified a larger list of DNT compounds that can be used as a reference set for developing and evaluating alternative test systems. A list of 66 compounds with different types of positive and negative controls, and respective comments on mode-ofaction was compiled specifically for DNT assay establishment. Amongst this list, only 10 toxicants fulfilled the stringent selection criterium of human evidence. A larger list of about 100 compounds was compiled as part of a published workshop report describing criteria to be applied in DNT test system establishment (Crofton et al., 2011). This list has been complemented with additional background information (e.g. reference to the respective animal studies) and re-published to support the development of high-throughput screening systems (Mundy et al., 2015). This extensive list contains both direct- and indirect-acting compounds, and the quality of the underlying publications shows a large variability. For the present study, a different approach was taken to assemble a list of reference compounds. The main goals were (i) to identify a practical number of chemicals for assay development (about 30 compounds),; (ii) to define clear selection criteria with regards to the published data, and the statistical methods applied to the data reported in these publications; (iii) to document failures to fulfill the selection criteria, and to communicate considerations concerning the use of this compound set for assay development. The intention was not to investigate all potential DNT compounds. For this process, a group of scientists assembled at a workshop developed an initial list of suggested compounds. During the follow-up period, four independent rounds of review by different subgroups of scientists with relevant expertise, resulted in a consensus set of 33 DNT test compounds.


Aschner et al.

Page 4

1.2 Adverse outcome pathways and fundamental neurobiological processes

Author Manuscript

Assays (see Box 1 for a glossary) for rapid screening of chemicals with a potential to cause DNT will likely use in vitro approaches or alternative models (Bal-Price et al., 2010; Coecke et al., 2007; Smirnova et al., 2014) that are compatible with high throughput screens. The feasibility and utility of such tests is based on the measurement of cellular perturbations relevant to neurodevelopment in humans (Bal-Price et al., 2015b; Kadereit et al., 2012; Lein et al., 2005). The predictive power of these assays will depend on the strength of association between the test endpoints assessed and the neurodevelopmental impairment observed in exposed human populations (or representative mammalian animal models).

Author Manuscript

In order to facilitate the development and use of molecular and cellular endpoints in predictive assays, the concept of the adverse outcome pathway (AOP) has recently been introduced (Ankley et al., 2010). AOPs are conceptual constructs that link a molecular initiating event (MIE) and an adverse outcome at the level of the whole organism (Tab. 1). A molecular initiating event is the initial point of contact between a chemical and a specific biomolecule that results in a cascade of key events (KE) leading to an adverse outcome (BalPrice et al., 2015b; Leist et al., 2014). For example, the binding of domoic acid to the glutamate receptor can result in a series of events that result in seizures and memory loss (Bal-Price et al., 2015b; Leist et al., 2014; Watanabe et al., 2011).


In the case of chemicals that cause DNT, most AOPs lack sufficient quantitative features (i.e. quantifiable key event relationships, such as activation thresholds and quantitative timeconcentration-effect relationships) to allow specific associations between the molecular initiating event and toxicity manifested at higher levels of biological organization. For this reason, it has been suggested that the first generation of new test methods for developmental neurotoxicity should focus on the assessment of a chemical’s ability to interfere with superordinate ‘fundamental neurodevelopmental processes’ (Lein et al., 2005; Bal-Price et al., 2015a). Studies on neurodevelopment in a variety of invertebrate, non-mammalian vertebrate and mammalian organisms (including man) indicate that the fundamental biological processes of neurodevelopment are remarkably conserved across species (Albright et al., 2000; Cowan et al., 1997; Thomas, 2001; Thor, 1995; Tropepe and Sive, 2003), even though small but distinct differences exist at the mechanistic level, especially the timing of events (Balmer et al., 2014; Smirnova et al., 2015). These ‘fundamental biological/neurodevelopmental processes’ include neural cell proliferation and differentiation, neuronal and glial cell migration, axonal and dendritic outgrowth as well as synapse formation and stabilization, apoptosis and myelination (Fig. 1) (Hoelting et al., 2015; Smirnova et al., 2015; van Thriel et al., 2012). Additional overarching processes, mostly limited to pathological situations reflect different states of glial activation, often termed neuroinflammation (Falsig et al., 2004; Kuegler et al., 2010; 2012; Zerrate et al., 2007). The final outcome of the tightly regulated spatiotemporal execution of these neurodevelopmental processes is the formation of functional signaling networks, and both experimental and clinical studies demonstrate that disruption of the spatiotemporal patterns or magnitude of any of these fundamental processes can significantly alter network connectivity and thus impair neural network function (Tab. 2) (Barone et al., 2000; BergerSweeney and Hohmann, 1997; Deoni et al., 2011; Deutsch et al., 2010; Gatto and Broadie,


Aschner et al.

Page 5

Author Manuscript

2010; Jones et al., 2000; Semrud-Clikeman and Ellison, 2009; Smirnova et al., 2015). Because cell-based assays that replicate these fundamental neurodevelopmental processes integrate effects across multiple molecular targets and mechanisms of action, and simple organism-based models additionally integrate effects across multiple cell types and organ systems, these alternative models can “cast a wide net” for detecting chemicals that act through diverse, and potentially unknown, molecular initiating events. Multiple such assays have been developed, e.g. using combinations of human neural cell types, or model organisms like zebra fish, and work with such methods is ongoing to clarify which of the perturbations that are observed show sufficient sensitivity and specificity to be used for predicitions of human adverse effects (Bal-Price et al., 2015b; Bal-Price et al., 2012; Crofton et al., 2011, 2012; Smirnova et al., 2014; van Thriel et al., 2012). 1.3 Linking of test systems and apical DNT endpoints


Adverse outcome pathways represent one of several concepts that have been developed to describe the chain of events that link exposure of a biological system to a xenobiotic with the hazard it poses. The concepts differ according to their focus on particular components within the chain of events, and on the intended use of the construct. Quantitative descriptions of the network of cellular events that decide the eventual cell fate are the focus of the ‘pathways-of-toxicity’ approach (Bouhifd et al., 2015; Hartung and McBride, 2011; Kleensang et al., 2014). In vitro toxicity testing is the major focus of the ‘biomarkers-oftoxicity’ concept, which concerns the identification of measurable and predictive endpoints that can be applied to model systems. For the purpose of compound selection for DNT in vitro assays, the concept of ‘toxicity endophenotypes’ contributes a useful perspective (Kadereit et al., 2012; Balmer and Leist, 2014; Bal-Price et al., 2015a) (Fig. 2). It focuses on fundamental biological processes of relevance to adverse outcomes at the organismal level that can be modeled by in vitro systems.

Author Manuscript

Characteristic adverse outcomes in the field of DNT are cognitive or psychomotor deficits, including reduced IQ, attention deficit, ataxia or various sensory disturbances, in addition to malformations (e.g. spina bifida or microcephaly). They describe external/apical phenotypes that are functionally defined, and which are difficult to model using presently-known in vitro systems. Unfortunately, most knowledge on human DNT compounds relates to these externally manifested functional phenotypes (= exophenotypes). For development of relevant model systems, we need approaches to link the ‘exophenotype’ caused by xenobiotic exposure in the intact organism to the effects the compound triggers in in vitro test systems. Such associations are the particular focus of the concept of toxicity endophenotypes. Endophenotypes are a description of the altered biological state of the nervous system in vivo that underlie the exophenotype. In less theoretical terms, ‘toxicity endophenotypes (TEP)’ describe the altered functional or structural connectivity or responsiveness of parts of the nervous system triggered by xenobiotics, and they represent the level of organization that links in vitro test systems for fundamental biological processes to apical DNT endpoints (exophenotypes). All developmental neurotoxicants are expected to affect at least one fundamental biological process in vivo, and this would result in an altered TEP. Thus TEP represent a key link between the known effects of DNT chemicals and their effects in in vitro systems. (see Tab. 2).


Aschner et al.

Page 6

Author Manuscript

The concept of TEP is also helpful for interpreting test results, evaluating their relevance and choosing endpoint-specific tool compounds in such systems. In this context, it is important to distinguish between the TEP (a state that is assessed in vivo) and the disturbed biological processes that led to it (and which may be assessed in vitro). For instance, a disarray of cells in a certain brain region may be the result of inhibited migration, altered patterning or even reduced neurite outgrowth that prevents axons from reaching appropriate target regions, and therefore results in apoptotic elimination or aberrant wiring. 1.4 Practical implications for the choice of positive-control compounds

Author Manuscript Author Manuscript Author Manuscript

The theoretical dissection of various associations relevant for the interpretation of DNT test system data (Exophenotype ←→ endophenotype ←→ biological processes ←→ test systems) has important practical significance, for instance by identifying research gaps and showing needs for further biological information. An important knowledge gap for DNT toxicants is the link between disturbed fundamental biological processes and TEP. This essential piece of information is difficult to obtain, as there is often a delay between chemical disturbance of a neurodevelopmental process and the DNT manifestation. Without knowledge on this link, it is not possible to define positive control toxicants for in vitro test systems that reflect only one of few biological processes relevant for DNT (Westerink, 2013). This has three important consequences. The first is that evaluation of test system performance (predictivity) with ‘known’ DNT chemicals is problematic using the standard approach of statistical correlation. The first type of misinterpretation are false negatives. If a test system does not react to a given DNT compound, the test system would be interpreted as lacking sensitivity, even though many DNT compounds would correctly show no effect in any given test system. In these cases, compounds cause their toxicity by affecting fundamental biological processes that are not captured by the test system in question. For instance, test systems that evaluate neurite extension or synapse formation would not be expected to react to methylazoxymethanol (MAM), an established DNT chemical (Penschuck et al., 2006) that affects precursor cell proliferation. A second type of misinterpretation/pitfall refers to examples of false positives that occur if a test system reacts to a compound that does not cause DNT in humans (in vivo) by altering the biological process evaluated in this system. For instance, if MAM, which as indicated above is a compound that specfically affects dividing cells, shows an effect in a test system of synapse formation, this would most likely be a false positive, from the point of view of mechanistic toxicology. However, it needs to be noted that it could be a true positive affecting a target different from DNA that has simply not yet been identified in in vivo systems due to their low sensitivity and high noise. Practical example for such a case are found when examining litereature on direct effects of chlorpyrifos on biological systems in vitro. For instance, voltage -gated calcium channels are inhibited by the parent compound, while the wellestablished inhibition of acetylcholine esterase is more sensitive to the oxon metabolite (Meijer et al., 2014a,b). The second consequence is that sets of compounds other than ‘gold standard DNT chemicals’ are required to initially evaluate the performance of in vitro test systems. Such chemicals should affect the known biology and mechanisms of the test system in defined, and, preferentially, specific ways. These compounds, here termed ‘endpoint-specific


Aschner et al.

Page 7

Author Manuscript

controls’ or ‘endpoint-specific reference compounds’ (Tab. 3), are in many cases not known to be associated with DNT. Therefore, the evaluation of the usefulness and relevance of the test system would not be possible through correlation of chemical’s effects in vitro vs. in vivo. It rather needs to be based on biological plausibility. One of the experimental approaches to this issue is the identification of the signaling processes governing the test system and their mechanistic relevance to signaling processes known to control the corresponding biological processes in vivo. The relevance and role of such signaling processes could be tested with sets of mechanistically-defined tool compounds. This would help to link the underlying biology of the test system to TEPs that are produced by genuine DNT compounds.

Author Manuscript

A third consequence is that the major usefulness of a set of positive DNT compounds lies in the establishment and evaluation of a test battery, rather than individual assays. The serious limitations that apply to individual tests (see first consequence) do not apply to a test battery that aims to cover the majority of DNT adverse effects. Compounds that are defined as gold standard positive controls should be identified as hits in the test battery (or an associated integrated approach to testing and assessment (IATA)). If they are not identified in the test battery, they would be correctly classified as false-negatives. Vice versa, negative controls should not be identified as hits, or they would be classified as false positives. Thus, a set of control compounds would be useful to evaluate an IATA approach (Bal-Price et al., 2015b; Rovida et al., 2015), and at the same time they would be useful in guiding the establishment of a test battery and for identifying data gaps to be filled using tests of higher sensitivity for specific compounds.

2 Endpoint-specific control compounds Author Manuscript

2.1 The concept of endpoint-specific control compounds

Author Manuscript

Assays (test methods) for DNT propose the use of both in vitro models based on neural cell cultures and alternative (non-mammalian) species as test systems. This guarantees that there will be a wide variety of measurements used to detect a change induced by a test substance, ranging from molecular (e.g. RNAs, proteins) to biochemical (e.g. neurotransmitters and their receptors) to morphological (e.g. cell size, shape or motility) to functional (e.g. locomotor activity, receptor function, electrophysiological properties). These measurements, regardless of the format, should assess an endpoint related to a fundamental neurodevelopmental process. A particular test system may allow for assessing multiple endpoints related to the same neurodevelopmental process. For example, the endpoint of proliferation can be assessed using both biochemical measurements of the amount of DNA and the morphometric assessment of cell numbers. As part of the setup and evaluation of a new test method, it should be demonstrated that measures for an endpoint are robust, reproducible (Miller, 2014; Poland et al., 2014) and accurate, and that the dynamic range within the test system is characterized. Moreover, different ways of measuring the same endpoint should yield similar results (consistency of readout). The next crucial step is the demonstration that a chemical-induced change in the biological endpoint can be detected. To describe this phase of assay evaluation, the concept of endpoint-specific controls has been introduced. Endpoint-specific controls (also termed ‘endpoint-selective controls’ or


Aschner et al.

Page 8

Author Manuscript

‘mechanistic tool compounds’) (Crofton et al., 2012; Kadereit et al., 2012; Leist et al., 2010) are chemicals that are known to reliably alter the endpoint of concern in a particular test system. Ideally, endpoint-specific control chemicals would be used to demonstrate both an increased and decreased response. They are selective in that within a known concentration range, the chemical will alter the primary test endpoint (e.g. precursor cell proliferation) without affecting general test system characteristics, including measures of cell viability. To continue with the example of proliferation, an endpoint-specific control would decrease (or increase) the measures of DNA and cell number within a test system in the absence of a change in cell viability. For neural cell proliferation, such chemicals include those with a known mechanism (e.g. the DNA polymerase inhibitor aphidicolin or the spindle poison taxol) or those where the mechanism is unclear but for which there is substantial literature evidence demonstrating selectivity (e.g. cadmium for certain systems).


Endpoint-specific controls are typically used in the initial evaluation of assay performance. In this sense, they are considered as “positive control” chemicals since they should be chosen based on prior knowledge that they alter the endpoint of concern under similar conditions using an established measurement. For example, studies from multiple laboratories have demonstrated that the MEK (MAP kinase kinase) inhibitor U0126 decreases neurite length in PC12 cells in a concentration-dependent manner (Kano et al., 2002; Liu et al., 2006). Thus, U0126 was used as an endpoint-specific control to determine if biochemical assessment of GAP-43 was a suitable measurement for neurite outgrowth in PC12 cells (Das et al., 2004). In the case where the test system is capable of producing an endpoint response in both directions, endpoint-specific controls for both an increased response and decreased response are desirable. For example, neurite outgrowth in PC12 cells can be increased above that measured under standard culture conditions by treatment with the IP3 kinase inhibitor C5 (Eva et al., 2012). Once an endpoint-specific control for a particular test system has been identified and characterized, it can be used as a “withinassay” or “within-plate” reference control during chemical testing. This internal control helps to identify plate-to-plate or test-to-test variability and to establish historical response levels. This is done by including one or more replicates containing a concentration of the endpoint-specific control known to produce a measurable response in the endpoint of interest without altering other outcomes. Moreover, such reference measurements can be used to define acceptability criteria for test results (on a per-plate or per-day basis). 2.2 Selection of endpoint-specific controls

Author Manuscript

The selection of endpoint-specific control compounds should be based both on the fundamental neurodevelopmental event being assessed and the test system being used. Prior knowledge of developmental neurobiology may identify signalling cascades required for the biological process evaluated in the test system and/or suggest pharmacological or drug-like chemicals that specifically target those signaling pathways. These “mechanistic tools” (Kadereit et al., 2012) would have a high probability of a positive effect within the context of a test system for a specific system’s endpoint. However, knowledge of the “mechanism” of a chemical is not a prerequisite for identifying an endpoint-specific control if there is sufficient evidence showing selective effects on an endpoint within a test system. The


Aschner et al.

Page 9

Author Manuscript

following criteria should be considered when identifying chemicals to be used as endpointspecific controls: Peer-reviewed data—Of primary importance is the previous demonstration in the peerreviewed literature that a chemical alters the endpoint within a particular test system. Reliability of the effect is demonstrated by showing the full concentration-response behaviour, providing evidence for the selectivity of the chemical for the endpoint of interest compared to other possible outcomes (e.g. cytotoxicity, metabolic competence, etc.). Demonstration of mechanistic consistency is highly desirable, e.g. demonstration that a kinase inhibitor indeed inhibits the target kinase in the relevant concentration range (in which it affects the systems endpoint) in the given test system. Studies using a single concentration or without a concurrent measure of general cell health do not provide sufficient data to identify endpoint-specific controls.

Author Manuscript

Demonstrated effects in multiple test systems—The demonstration that a chemical meets the criteria listed in criterion A (above) in more than one test system (e.g. different cell types) or under multiple conditions (e.g. different cell culture media or different periods of exposure) increases confidence in its application as an endpoint-specific control. Data for the same chemical should ideally be available from multiple laboratories.

Author Manuscript

Knowledge of chemical mechanisms—Chemicals with a known target (molecular initiating event) or known actions at various levels of biological organization increase reliability for a selective effect on a particular neurodevelopmental endpoint. Knowledge of the signaling pathways underlying a fundamental neurobiological process in a given test system can help to identify potential endpoint-specific controls. Sometimes test system development will require acquisition of this biological knowledge, by screening of known pathways or identification of new pathways by broad screening approaches and use of omics methods.

Author Manuscript

Chemical causes same qualitative effect in vivo—Some endpoint-specific controls may cause the same qualitative effect in an in vitro test system and in vivo, i.e. it may affect the fundamental neurodevelopmental process that is modelled in the in vitro test in a live developing mammal. The congruence of results from standard (in vivo) and alternative test methods (in vitro/lower model organisms) increases confidence that the chemical is selectively acting on a fundamental neurodevelopmental endpoint. However, this is not a mandatory criterion, as several good endpoint-specific controls may not be active in vivo, due to metabolism, toxicokinetic reasons or off-target toxicity. Based on these criteria, endpoint-specific control compounds for fundamental neurodevelopmental processes have been compiled (Tab. 4). 2.3 Selection of negative controls Once an assay has been established and has been shown to react to endpoint-specific controls, some basic evaluation of specificity is important. This requires compounds that have no effect in the test system. Such negative controls do not perturb the respective fundamental neurodevelopmental process, or its underlying signalling pathways. The ideal


Aschner et al.

Page 10

Author Manuscript

negative controls can be defined as chemicals that are biologically (pharmacologically) active in other systems, but are not expected to have an effect on the endpoints of the test system under evaluation. To demonstrate absence of effect, a concentration should be used that shows a significant effect in other test systems.

Author Manuscript

In practice, it is sometimes difficult to identify pharmacologically potent compounds devoid of any DNT effect. In such cases, the simplest type of negative controls are compounds that do not cross the cell membrane (such as mannitol). Groups of chemicals with good potential as negative controls are nutrients (e.g. ascorbic acid), chemicals that target other organ systems (e.g. the liver toxicant paracetamol), or chemicals with a known target (molecular initiating event) that is not expressed in the test system (e.g. the proton pump inhibitor omeprazol) (Kadereit et al., 2012). Alternatively, drugs that are recommended for use in pregnancy are an important resource, but all of them require individual evaluation. Few suggestions for negative controls for evaluation of DNT assays have been compiled (Tab. 4). For these compounds, no peer-reviewed papers reporting on their developmental neurotoxicity could be identified. Preference is given to compounds that have been actively tested for DNT, but were found experimentally to be negative. 2.4 How to deal with specificity

Author Manuscript

Many published test systems reach high levels of sensitivity for some known DNT compounds, but little information is available on specificity. This issue is directly related to the topic of compound selection for DNT test systems, as specificity of a test system is defined as the capacity to classify negatives correctly, i.e. specificity correlates with a low rate of false positives. Thus, selection and testing of negatives is an essential step in the optimization cycles of test system establishment. This task is not trivial, as it is not sufficient to simply select compounds for which there is currently no evidence that they trigger DNT.

Author Manuscript

Three considerations are important for the selection of good negative controls for specificity testing: (i) First, the biological process modeled in a test system is not the same as the phenotype resulting from exposure to a DNT chemical in vivo (see TEP above). Therefore, ‘non-DNT chemicals’ may specifically affect a test system (see endpoint-specific controls above), and the task to find real negatives is often difficult, and it needs to be determined for each test system; (ii) The second reason is the potential for interaction of test endpoints. For instance, viability and neurite growth are two endpoints in a given test system, but they are not independent of one another. For example, some xenobiotics may affect a specific test endpoint (neurite growth) indirectly by acting on cell viability. Thus, such compounds would appear as positive hits, although they are true negatives with respect to the primary biological process (neurite growth) examined in the test system. The most frequent of these phenomena is decreased cell viability by a ‘nonspecific’ test compound, which subsequently influences the test endpoint(s) of primary interest. Therefore, care needs to be taken that overall reduced cell viability or decreased cell survival is not interpreted as an effect on differentiation, neurite growth, migration or synaptic connectivity (all of which may also be affected because viability is reduced). A straightforward approach to this problem is testing of compounds only at concentrations determined to not cause cytotoxicity in that test system. However, unambiguous definitions on how non-cytotoxic concentrations should be


Aschner et al.

Page 11


determined do not exist at present. To eassess the specificity of a test system for directacting DNT compounds, it is necessary to select a second group of negative control compounds, i.e. nonspecific controls known for their general cytotoxicity (Kadereit et al., 2012; Leist et al., 2010). The concentration ratio of these compounds concerning specific (e.g. neurite growth) and nonspecific (e.g. cytotoxicity) test endpoints can be used to define a prediction model for test specificity (Krug et al., 2014; Stiegler et al., 2011); (iii) The third problem is related to toxicokinetics (including drug metabolism). Several compounds would (based on their biochemical activity) affect fundamental neurodevelopmental/biological processes relevant to DNT, but they are not recognized as DNT compounds in the literature or by in vivo testing, as they do not reach the fetus or the central nervous system at the doses used. Such compounds would be scored as false positives in in vitro assays, with respect to in vivo effects, but they would in fact be true positives with respect to the biology tested in the assay. Thus, a task for the future would be to provide background (toxicokinetic) information on such effects and compounds.

3 Selection of high-quality DNT reference compounds 3.1 Selection procedure and rules

Author Manuscript

A group of neurotoxicology experts from government, academia and industry convened in Konstanz, Germany, (October, 2011) to identify chemicals for potential use as positive controls for developmental neurotoxicity. The selection was based on two major principles: (a) the list of chemicals was intended to be exemplary, and not exhaustive. The initial selection of candidates did not follow a defined screening process or data base search algorithm, rather it was based on the subjective recall of the experts of frequently-quoted litereature or their own work. The aim was to establish a list of 20–30 compounds useful for assay development and evaluation, and compounds with solid evidence for DNT activity may not have been considered; (b) after compilation of a primary list, compounds were vetted using pre-defined criteria (Box 2). The purpose of the selection criteria was to ensure that the selection process was based on scientifically sound studies. Moreover, the goal was to increase the likelihood that the selected positive controls act as direct developmental neurotoxicants, and that adverse effects are not the indirect consequence of maternal toxicity. The supplementary material contains extensive information on the low-effect-levels (LOELs) and no-effect-levels (NOELs) for offspring, maternal toxicity and the DNT endpoints affected.

Author Manuscript

Candidate compounds that largely failed to meet these criteria were eliminated from the list. Compounds that met many of the criteria were retained, and the criteria that were not met are flagged. In general, the supporting documentation for these compounds derives from published animal studies, but in some cases, human epidemiological evidence based on multiple studies was available as additional supportive evidence. Most of the evidence on human effects is derived from authoritative reviews (Grandjean and Landrigan, 2006, 2014) that compiled available evidence for DNT effects in a systematic way. However, complete weight-of-evidence evaluations are available for all compounds. For example, there is still controversy in the field as to the relevance of DNT effects of chlorpyrifos at human exposure levels (Burns et al., 2013; Li et al., 2012; Mount et al., 2009).


Aschner et al.

Page 12

Author Manuscript

The list of DNT reference chemicals (Table 5 and supplementary material) should be considered a sample list of positive control chemicals that have the potential of causing developmental neurotoxic effects in animals at some dose level, which may or may not be relevant to human exposure levels. Of the 33 compounds listed, the majority (n = 29) overlap with the more extensive list assembled by scientists from the EPA (Mundy et al., 2015). The non-overlapping references suggested here are the pesticide lindane, the recreational drug 3,4-methylenedioxy-N-methamphetamine, and the groups of perfluorinated aliphatic compounds comprising perfluoro-octanoic acid (PFOA) and perfluoroactane-sulfonic acid (PFOS).


Note, the list of reference DNT compounds presented here requires an evaluation of its fitfor-purpose by end-users, and this implies elimination or addition of compounds for specific purposes or additional literature searches on specific compounds within the list. A future step may be the compilation of systematic reviews on each of the compounds, with respect to the weight of evidence that they are developmental neurotoxicants in animals. For instance, here only positive evidence for DNT effects was considered. It was neither weighed against the entirety of the available literature of a given compound (which may also include negative studies), nor did we consider that there may be a publication bias (with negative findings less likely to be published). A systematic review would also provide information on whether a parent compound acts directly as developmental neurotoxicant as well as the role of metabolism in toxifying or detoxifying the parent compound. This consideration is pivotal for chemical use in in vitro systems as well as alternative species models in which metabolism can vary from that of humans. For instance, chlorpyrifos may need to be converted to chlorpyrifos-oxon (Yang et al., 2008), heroin may fail to show effects in systems that lack deacetylases that catalyze the formation of the final toxicant morphine, and 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) will fail to show any effect, unless it is metabolized by astrocytic monoamine oxidase to 1-methyl-4phenylpyridinium (MPP+) (Efremova et al., 2015; Schildknecht et al., 2015). 3.2 Use of the DNT compound set

Author Manuscript

After setup of a test, and evaluation of its technical performance and reproducibility on the basis of endpoint-specific controls, the next steps involve, amongst others: (i) gathering information on the predictivity of the test; (ii) establishment of a prediction model; (iii) introducing additional endpoints and/or adjusting parameters for increased rate of data collection or higher throughput; (iv) identification of biomarkers, measurable key events or signaling mechanisms that may be used to build or optimize other test systems, or translation to human studies; and (v) characterization of the MOA of known DNT toxicants to evaluate which AOP key events are reflected by the test system, and for which types of mechanisms the test is applicable. For such activities, a set of highly relevant (i.e. high confidence that they indeed trigger DNT in vivo) control compounds is essential. For instance, one traditional way to evaluate predictivity would be to split the pool of DNT compounds into a training set and a testing set. Using the data generated with the training set plus negative controls, a prediction model would be established. The validity of this model, and its performance (accuracy, specificity, sensitivity) would then be tested by blinded


Aschner et al.

Page 13

Author Manuscript

measurement of the testing set. In a variation of this approach, the splitting of the compounds into training and test sets would be done in silico in many, or in all combinatorily possible ways, after all compounds have been tested. Introduction of new endpoints or identification of biomarkers (Krug et al., 2014; Stiegler et al., 2011; Zimmer et al., 2012) requires the availability of a relevant set of test compounds that allow correlation studies from one system or from one endpoint to another. To an even greater extent, the same holds true for identifications of general toxicity mechanisms (Fritsche et al., 2005; Gassmann et al., 2010, 2014; Langeveld et al., 2012; Lein et al., 2007; Waldmann et al., 2014; Yang et al., 2014; Zimmer et al., 2011a) or for development of toxicant classifiers (Krug et al., 2013b; Rempel et al., 2015), as the selected compounds are the main anchoring point of such studies.

Author Manuscript

4 Challenges encountered during the search for reference compounds 4.1 Research bias Two examples have been selected here (phenytoin, isotretionin, see sub-chapter below) to illustrate the challenges of selecting reference compounds for DNT, based on criteria of high quality data, and statistically sound human or animal studies.

Author Manuscript

Concerning animal data, the studies are often old, and the design and reporting standards are not up to current demands for documenting a gold standard reference compound. Some studies only show a (non-significant) trend or a possibility that a compound is a DNT toxicant. Nevertheless, such data may have important implications for further handling of such compounds. Such initial findings may prevented further studies to establish statistical significance of the effects and to meet the quality standards established here for compound selection. This may have been due to several reasons. For instance, institutional or regulatory approval for animal experimentation are hard to obtain if an experiment is mainly confirmatory of previous findings (even though these are not of high quality). Another reason is that funding is hard to obtain for confirmatory studies that differ from earlier findings mainly in statistical power and rigor of design.

Author Manuscript

Concerning human data, a similar situation is frequently observed, i.e. initial weak evidence makes it difficult to obtain further, more definite evidence. The major reason here is that once a potential hazard has been documented, measures will be taken to reduce the risk, i.e. human exposure to the compound in question is kept to a minimum. Therefore, obtaining epidemiological data on compounds with a suspected DNT hazard is particularly difficult. A way around the problems described above could be the increased use of a battery of alternative methods that is sufficiently evaluated for its performance and predictivity. 4.2 Phenytoin and isotretinoin exemplify challenges in obtaining high quality literature data This situation is demonstrated by two suspected DNT compounds, phenytoin and isotretinoin. They did not fully fulfill the statistical and documentation criteria identified in Box 2, but they were included (see details below) in our compound collection (Tab. 5) with indication of the limitations of the available published literature.


Aschner et al.

Page 14

Author Manuscript

Diphenylhydantoin (phenytoin) is a sodium channel blocker used as an anticonvulsant antiepileptic drug. In the literature, a malformation, called ‘fetal hydantoin syndrome’ is observed in children exposed to phenytoin during fetal development. Fetal hydantoin syndrome is associated with cerebellar malformations and psychomotor dysfunction after intrauterine exposure (extensively reviewed by Vorhees (Vorhees, 1994)). Several animal studies are suggestive of hydantoin being a DNT toxicant. Described effects range from impaired synapse function (Forcelli et al., 2012) and neurodegeneration (Asimiadou et al., 2005) to general neurotoxicity (Hatta et al., 1999). However, the studies fail to fulfill the full set of criteria, defined by the workshop participants for a DNT reference compound (Box 2, Tab. 5). There are also several reports that suggest phenytoin is a human DNT toxicant, but a review (Nicolai et al., 2008) covering 56 studies concerning teratogenic effects of antiepileptic drugs, concluded: “The identified studies do not allow definite conclusions. The

Author Manuscript

possibility of neurodevelopmental delay, behavioural disorders, or learning disabilities as an outcome of in utero exposure to AEDs needs to be considered seriously. The literature however does not provide evidence for a valid risk estimate”. Isotretinoin is one of the isoforms of retinoic acid (usually the generic name retinoic acid refers to the all-trans isoform, while isotretinoin has one cis-bond (position 13). It is the active ingredient in the highly effective antiacne drug Accutane and is suspected to cause depression and suicide in adults and neonatal malformations. From 1982 to 2006, more than 2,000 isotretinoin users became pregnant. Amongst them, a high frequency of spontaneous or elective abortions was observed. As of 2002 — the year generic Accutane was approved — the FDA had received reports of 172 babies born with a congenital defect or anomaly after maternal use of Accutane2,3. They quote: “Accutane is clearly a potent human


teratogen that causes malformation of the central nervous system, cardiovascular system and facial structures”. This is, however, not supported by animal studies that meet the quality criteria set out here. The reason is interesting and very instructive. Already in the 90s it became clear that the teratogenicity of some compounds depends on pharmacokinetics (Nau, 1986). Isotretinoin (Nau, 2001) is one of the drugs that shows negligible effects in mouse and rat (Kochhar and Penner, 1987; Kamm, 1982), while in monkeys (Fantel et al., 1977) and (possibly) humans, isoretinoin can cause great disturbances of embryonic development. It is assumed that most effects of isotretinoin (13-cis-retinoic acid) are mediated by isomerization to all-trans-retinoic acid. Concerning this metabolic prerequisite, the situation has been described as follows: ‘The insensitive species (rat, mouse) eliminate the drug rapidly through detoxification to β-glucuronide; also, placental transfer is limited in these species. On the other hand, in sensitive species (primates), the drug is predominantly metabolized to the active 13-cis-4-oxo-retinoic acid; placental transfer is more extensive here’ (Nau, 1986). The two above examples clearly demonstrate the difficulties with compiling a definite and exhaustive list of DNT chemicals. Likely there are other compounds that could be included in the list, and, there are likely many compounds that are DNT toxicants, but that lack

2http://www.drugwatch.com/accutane/side-effects.php 3http://www.cdc.gov/mmwr/preview/mmwrhtml/00000310.htm


Aschner et al.

Page 15

Author Manuscript

sufficient animal or human data to be considered gold standard reference compounds for test evaluation. 4.3 Examples of other compounds not considered here

Author Manuscript

The test set presented here may be complemented by additional compounds as determined by personal preference or scientific needs. They may be selected from a recently-published 100 compound collection or from newly emerging publications on DNT (Mundy et al., 2015). In all cases, it is advisable to apply the criteria delineated in Box 2 to additional compounds. Amongst the more reently discussed copounds with a potential to cause DNT is paracetamol (Brandlistuen et al., 2013; Liew et al., 2014; Viberg et al., 2014), but it is not clear yet whether this effect is direct or whether it requires metabolic activation. There are also indictions that the food-borne non-proteinogenic amino acid BMAA affects neurodevelopment (Karlsson et al., 2015). The same is true for acrylamide, a chemical generated from amino acid precursors during food processing (Duarte-Salles et al., 2013; Pedersen et al., 2012). However, more information regarding specificity is required; for example, acrylamide’s effects on head circumference and brain weight may also be indirect consequences of toxicity. Also not included here is the developmental toxicant cyclopamine (Cooper et al., 1998), a plant ingredient with broad developmental effects that is listed amongst the endpoint-specific controls for neurodifferentiation assays.

5 The path forward 5.1 How to get more mechanistic information on DNT compounds?

Author Manuscript

One of the major problems for developing and evaluating DNT assays remains the fact that there is a paucity of information regarding the effects of DNT compounds on fundamental neurobiological processes in humans. This precludes an evaluation of test predictivity based solely on the correlation of its results with in vivo findings (Leist et al., 2012). One way forward would involve the three following activities: (a) obtaining more knowledge on modes of action of DNT chemicals by profiling them in a broad set of well characterized and robust in vitro test systems (Behl et al., 2015; Zimmer et al., 2014; Daneshian et al., 2016; Hirsch et al., 2016; Pallocca et al., 2016); (b) optimizing in vitro test systems, by using endpoint specific controls and already well-characterized DNT compounds; (c) using steps (a) and (b) in an iterative fashion to optimize test systems and test methods.

Author Manuscript

The path forward also involves increased greater understanding of the biology underlying the test systems, understanding why certain compounds work or do not work, and learning exactly why DNT reference compounds work in some systems, but not in others. This process requires mechanistic interventions, follow-up on pathways-of-toxicity and studies of groups of related compounds (Dreser et al., 2015; Krug et al., 2013a; Krug et al., 2013b; Zimmer et al., 2011a). Most likely, test systems will need to be characterized by many different analytical approaches to derive the needed information. Limitation to a single, toxicologically-relevant endpoint will not be sufficient in this establishment and optimization phase of a test system.


Aschner et al.

Page 16

5.2 How to deal with adversity vs adaptation


For all in vitro assays, it is difficult to distinguish between changes that are linked to adverse effects in vivo, and alterations that are adaptive or counter-regulators (Blaauboer et al., 2012). An overall solution to this challenge will be a major issue for the future. In the context of compound selection, a few points deserve immediate attention and action. The first and foremost is ‘concentration’. The questions of specificity and adversity cannot be linked to compounds as such, but only to a ‘compound at a given concentration’ (Waldmann et al., 2014; Daston et al., 2014). Although this appears trivial, it has hitherto been scarcely considered when specificity and sensitivity of an assay have been evaluated. In addition, most screens have up to now been performed at fixed compound concentrations that are not related to the pharmacological potency of the compounds screened. A change of this practice has been suggested for the ESNATS test battery (Pallocca et al., 2015; Zimmer et al., 2014), for which initial concentrations for testing have been based on a biological/mechanistic rationale. In addition, for many omics studies the chosen concentration is anchored to a biological effect (e.g. maximum non-cytotoxic concentration). In practice, the task of determining which concentrations are meaningful and correspond to in vivo effects is not trivial, and they can be quite difficult to determine (Westerink, 2013). A future useful step for the field would be the drafting of a consensus document addressing the feasibility of basing concentrations for DNT testing on reverse pharmacokinetic modelling (Bosgra et al., 2014). One of the approaches for defining adversity would be based on measuring concentration-dependency of many endpoints in the system and relating these dependencies to the concentration known to be associated with adverse effects in vivo. Another useful approach would be to not only rely on measurements at a defined time point at the end of the incubation, but to follow the temporal evolution of changes in the system in the absence versus presence of test compounds (Dreser et al., 2015).

Author Manuscript

5.3 How do we link test systems in vitro to DNT in vivo?

Author Manuscript

The usual evaluation of a test system addresses three domains: reproducibility, biological relevance and correlation with in vivo data (= predictivity). Determination of predictivity is only possible to a limited extent because of the lack of large numbers of well-characterized DNT chemicals, thus, more focus will need to be put on the first two domains (Basketter et al., 2012; Leist et al., 2012). A significant problem with the existing in vitro test systems for the identification of developmental neurotoxicants is the lack of explicit guidance on how to standardize DNT endpoints. Clear quality control procedures would be required for in vitro models to produce results comparable across laboratories, and with the ultimate goal to use data for regulatory purposes. To address biological relevance, several different approaches may be combined (Alepee et al., 2014; Hartung et al., 2013; Smirnova et al., 2015; van Vliet et al., 2014). One approach is directly related to the selection of test compounds: the understanding of the response to tool compounds, and mechanistically consistent responses to chemically-related compounds would be helpful to evaluate the biological relevance of the test system. Similar types of information for in vivo DNT, including information on the temporal evolution of the damage, would be very helpful.


Aschner et al.

Page 17

Author Manuscript

5.4 How can the information obtained using DNT reference compounds be applied to develop more predictive assays? The selection of chemicals that can serve as endpoint-specific controls will facilitate quality control and standardization of in vitro models. Systems would be expected to react in a predictable manner to positive and negative controls before they can be used further for chemical testing. Moreover, the study of DNT reference compounds compiled here will create an important data base for the characterization of new test systems, and for elucidating whether the ‘molecular machinery’ present in a cell system is capable of responding to known developmental neurotoxicants as expected.

Author Manuscript

The understanding of the pathways-of-toxicity/AOP induced by DNT reference chemicals could serve as a template to design assays that will be based on the key events that determine outcome. Such assays may have reduced complexity and higher throughput, and they would directly address selected AOP of relevance for DNT. To apply the AOP concept to DNT evaluation, a clear description of the measureable parameters is required to study each key event (Bal-Price et al., 2015a, 2015b; Edwards et al., 2016; Perkins et al., 2015; Tollefsen et al., 2014). With respect to the selection of chemicals and their characterization in DNT in vitro test systems, applying the AOP concept will provide important information for the development of structure-activity relationships (SAR) and “read-across”, i.e., using information from one chemical to predict the effects for another one, that is structurally related. This will allow grouping and ranking of chemicals according to their modes of action and potency (Dreser et al., 2015; Ramirez et al., 2013).

Author Manuscript

Based on comparing data generated across multiple diverse test systems, the most sensitive endpoints and the most reliable test systems could be selected for a ‘test battery’ as the basis for an IATA (see Box 1). One of the steps forward in this direction would be establishment of high-throughput screening assays. The data from such assays could be used for chemical prioritization, screening of chemicals for further in vivo testing (Bal-Price et al., 2012; Crofton et al., 2012, 2014; Judson et al., 2014), obtaining information on mixtures of compounds, integration of the data by systems toxicology methods (Hartung and McBride, 2011; Sauer et al., 2015), and reducing reliance on in vivo testing for regulatory decisionmaking.

Supplementary Material Refer to Web version on PubMed Central for supplementary material.

Author Manuscript

Acknowledgments This is a report of t4 – (transatlantic think tank for toxicology) series, sponsored by the Doerenkamp-Zbinden Foundation (DZF). The workshop was organized and sponsored by CAAT-Europe and DZF. The work of the authors was supported by various grants and home institution funds. They are grateful to EU-ToxRisk, the BMBF and the DFG (RTG1331, KoRS-CB, Memosystems). MA was supported in part by grants from the National Institutes of Health (R01 ES07331, R01 ES10563 and R01 ES020852). PL was supported in part by grants from the National Institutes of Health (P01 ES011269 and R01 ES014901) and the United States Environmental Protection Agency (RD835550 and RD83543201).


Aschner et al.

Page 18

Author Manuscript

References


Albright TD, Jessell TM, Kandel ER, et al. Neural science: a century of progress and the mysteries that remain. Neuron. 2000; 25(Suppl):S1–55. http://dx.doi.org/10.1016/s0896-6273(00)80912-5. [PubMed: 10718192] Alepee N, Bahinski A, Daneshian M, et al. State-of-the-art of 3D cultures (organs-on-a-chip) in safety testing and pathophysiology. ALTEX. 2014; 31:441–477. http://dx.doi.org/http://dx.doi.org/ 10.14573/altex1406111. [PubMed: 25027500] Ali MM, Murthy RC, Chandra SV. Developmental and longterm neurobehavioral toxicity of low level in-utero cadmium exposure in rats. Neurobehav Toxicol Teratol. 1986; 8:463–468. [PubMed: 3785508] Ankley GT, Bennett RS, Erickson RJ, et al. Adverse outcome pathways: a conceptual framework to support ecotoxicology research and risk assessment. Environ Toxicol Chem. 2010; 29:730–741. http://dx.doi.org/10.1002/etc.34. [PubMed: 20821501] Asimiadou S, Bittigau P, Felderhoff-Mueser U, et al. Protection with estradiol in developmental models of apoptotic neurodegeneration. Ann Neurol. 2005; 58:266–276. http://dx.doi.org/10.1002/ ana.20553. [PubMed: 16049923] Aziz MH, Agrawal AK, Adhami VM, et al. Methanol-induced neurotoxicity in pups exposed during lactation through mother: role of folic acid. Neurotoxicol Teratol. 2002; 24:519–527. http:// dx.doi.org/10.1016/S0892-0362(02)00231-3. [PubMed: 12127898] Bal-Price A, Crofton KM, Leist M, et al. International STakeholder NETwork (ISTNET): creating a developmental neurotoxicity (DNT) testing road map for regulatory purposes. Arch Toxicol. 2015a; 89:269–287. http://dx.doi.org/10.1007/s00204-015-1464-2. [PubMed: 25618548] Bal-Price A, Crofton KM, Sachana M, et al. Putative adverse outcome pathways relevant to neurotoxicity. Crit Rev Toxicol. 2015b; 45:83–91. http://dx.doi.org/ 10.3109/10408444.2014.981331. [PubMed: 25605028] Bal-Price AK, Hogberg HT, Buzanska L, et al. In vitro developmental neurotoxicity (DNT) testing: relevant models and endpoints. Neurotoxicology. 2010; 31:545–554. http://dx.doi.org/10.1016/ j.neuro.2009.11.006. [PubMed: 19969020] Bal-Price AK, Coecke S, Costa L, et al. Advancing the science of developmental neurotoxicity (DNT): testing for better safety evaluation. ALTEX. 2012; 29:202–215. http://dx.doi.org/10.14573/altex. 2012.2.202. [PubMed: 22892558] Balmer NV, Weng MK, Zimmer B, et al. Epigenetic changes and disturbed neural development in a human embryonic stem cell-based model relating to the fetal valproate syndrome. Hum Mol Genet. 2012; 21:4104–4114. http://dx.doi.org/10.1093/hmg/dds239. [PubMed: 22723015] Balmer NV, Klima S, Rempel E, et al. From transient transcriptome responses to disturbed neurodevelopment: role of histone acetylation and methylation as epigenetic switch between reversible and irreversible drug effects. Arch Toxicol. 2014; 88:1451–1468. http://dx.doi.org/ 10.1007/s00204-014-1279-6. [PubMed: 24935251] Balmer NV, Leist M. Epigenetics and transcriptomics to detect adverse drug effects in model systems of human development. Basic Clin Pharmacol Toxicol. 2014; 115:59–68. http://dx.doi.org/ 10.1111/bcpt.12203. [PubMed: 24476462] Baranski B. Behavioral alterations in offspring of female rats repeatedly exposed to cadmium oxide by inhalation. Toxicol Lett. 1984; 22:53–61. http://dx.doi.org/10.1016/0378-4274(84)90045-6. [PubMed: 6464034] Barone S Jr, Das KP, Lassiter TL, et al. Vulnerable processes of nervous system development: a review of markers and methods. Neurotoxicology. 2000; 21:15–36. [PubMed: 10794382] Basketter DA, Clewell H, Kimber I, et al. A roadmap for the development of alternative (non-animal) methods for systemic toxicity testing – t4 report*. ALTEX. 2012; 29:3–91. http://dx.doi.org/ 10.14573/altex.2012.1.003. [PubMed: 22307314] Behl M, Hsieh JH, Shafer TJ, et al. Use of alternative assays to identify and prioritize organophosphorus flame retardants for potential developmental and neurotoxicity. Neurotoxicol Teratol. 2015; 52:181–193. http://dx.doi.org/10.1016/j.ntt.2015.09.003. [PubMed: 26386178]


Aschner et al.

Page 19

Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Benesova O, Tejkalova H, Kristofikova Z, et al. Neuro-immuno-teratogenicity of drugs used in neonatal pharmacotherapy in relation to the ontogenic stage at the time of their administration. Gen Physiol Biophys. 1999; 18(Spec No):21–27. [PubMed: 10703715] Berger-Sweeney J, Hohmann CF. Behavioral consequences of abnormal cortical development: insights into developmental disabilities. Behav Brain Res. 1997; 86:121–142. http://dx.doi.org/10.1016/ S0166-4328(96)02251-6. [PubMed: 9134147] Blaauboer BJ, Boekelheide K, Clewell HJ, et al. The use of biomarkers of toxicity for integrating in vitro hazard estimates into risk assessment for humans. ALTEX. 2012; 29:411–425. http:// dx.doi.org/10.14573/altex.2012.4.411. [PubMed: 23138511] Bosgra S, van de Steeg E, Vlaming ML, et al. Predicting carrier-mediated hepatic disposition of rosuvastatin in man by scaling from individual transfected cell-lines in vitro using absolute transporter protein quantification and PBPK modeling. Eur J Pharm Sci. 2014; 65:156–166. http:// dx.doi.org/10.1016/j.ejps.2014.09.007. [PubMed: 25261337] Bouhifd M, Andersen ME, Baghdikian C, et al. The human toxome project. ALTEX. 2015; 32:112– 124. http://dx.doi.org/http://dx.doi.org/10.14573/altex.1502091. [PubMed: 25742299] Brandlistuen RE, Ystrom E, Nulman I, et al. Prenatal paracetamol exposure and child neurodevelopment: a sibling-controlled cohort study. Int J Epidemiol. 2013; 42:1702–1713. http:// dx.doi.org/10.1093/ije/dyt183. [PubMed: 24163279] Breier JM, Radio NM, Mundy WR, et al. Development of a high-throughput screening assay for chemical effects on proliferation and viability of immortalized human neural progenitor cells. Toxicol Sci. 2008; 105:119–133. http://dx.doi.org/10.1093/toxsci/kfn115. [PubMed: 18550602] Broening HW, Morford LL, Inman-Wood SL, et al. 3,4-methylenedioxymethamphetamine (ecstasy)induced learning and memory impairments depend on the age of exposure during early development. J Neurosci. 2001; 21:3228–3235. [PubMed: 11312307] Brys I, Pupe S, Bizarro L. Attention, locomotor activity and developmental milestones in rats prenatally exposed to ethanol. Int J Dev Neurosci. 2014; 38:161–168. http://dx.doi.org/10.1016/ j.ijdevneu.2014.08.007. [PubMed: 25192749] Burdan F. Intrauterine growth retardation and lack of teratogenic effects of prenatal exposure to the combination of paracetamol and caffeine in Wistar rats. Reprod Toxicol. 2003; 17:51–58. http:// dx.doi.org/10.1016/S0890-6238(02)00097-7. [PubMed: 12507658] Burns CJ, McIntosh LJ, Mink PJ, et al. Pesticide exposure and neurodevelopmental outcomes: review of the epidemiologic and animal studies. J Toxicol Environ Health B Crit Rev. 2013; 16:127–283. http://dx.doi.org/10.1080/10937404.2013.783383. [PubMed: 23777200] Burry M, Guizzetti M, Oberdoerster J, et al. Developmental neurotoxicity of toluene: in vivo and in vitro effects on astroglial cells. Dev Neurosci. 2003; 25:14–19. http://dx.doi.org/71463. [PubMed: 12876426] Butenhoff JL, Ehresman DJ, Chang SC, et al. Gestational and lactational exposure to potassium perfluorooctanesulfonate (K+PFOS) in rats: developmental neurotoxicity. Reprod Toxicol. 2009; 27:319–330. http://dx.doi.org/10.1016/j.reprotox.2008.12.010. [PubMed: 19162172] Cattabeni F, Abbracchio MP, Cimino M, et al. Methylazoxymethanol-induced microencephaly: persistent increase of cortical somatostatin-like immunoreactivity. Brain Res Dev Brain Res. 1989; 47:156–159. http://dx.doi.org/10.1016/0165-3806(89)90120-X. [PubMed: 2736763] Chattopadhyay S, Bhaumik S, Nag Chaudhury A, et al. Arsenic induced changes in growth development and apoptosis in neonatal and adult brain cells in vivo and in tissue culture. Toxicol Lett. 2002; 128:73–84. http://dx.doi.org/10.1016/S0378-4274(01)00535-5. [PubMed: 11869819] Coecke S, Goldberg AM, Allen S, et al. Workgroup report: incorporating in vitro alternative methods for developmental neurotoxicity into international hazard and risk assessment strategies. Environ Health Perspect. 2007; 115:924–931. http://dx.doi.org/10.1289/ehp.9427. [PubMed: 17589601] Coluccia A, Belfiore D, Bizzoca A, et al. Gestational all-trans retinoic acid treatment in the rat: neurofunctional changes and cerebellar phenotype. Neurotoxicol Teratol. 2008; 30:395–403. http:// dx.doi.org/10.1016/j.ntt.2008.03.064. [PubMed: 18495421] Cooper MK, Porter JA, Young KE, et al. Teratogen-mediated inhibition of target tissue response to Shh signaling. Science. 1998; 280:1603–1607. http://dx.doi.org/10.1126/science.280.5369.1603. [PubMed: 9616123]


Aschner et al.

Page 20


Cowan, WM., Jessell, TM., Zipursky, SL. Molecular and Cellular Approaches to Neural Development. New York: Oxford University Press; 1997. Vol Crofton K, Fritsche E, Ylikomi T, et al. International STakeholder NETwork (ISTNET) for creating a developmental neurotoxicity testing (DNT) roadmap for regulatory purposes. ALTEX. 2014; 31:223–224. http://dx.doi.org/http://dx.doi.org/10.14573/altex.1402121. [PubMed: 24794006] Crofton KM, Mundy WR, Lein PJ, et al. Developmental neurotoxicity testing: recommendations for developing alternative methods for the screening and prioritization of chemicals. ALTEX. 2011; 28:9–15. [PubMed: 21311847] Crofton KM, Mundy WR, Shafer TJ. Developmental neurotoxicity testing: a path forward. Congenit Anom (Kyoto). 2012; 52:140–146. http://dx.doi.org/10.1111/j.1741-4520.2012.00377.x. [PubMed: 22925214] Culbreth ME, Harrill JA, Freudenrich TM, et al. Comparison of chemical-induced changes in proliferation and apoptosis in human and mouse neuroprogenitor cells. Neurotoxicology. 2012; 33:1499–1510. http://dx.doi.org/10.1016/j.neuro.2012.05.012. [PubMed: 22634143] Dakshinamurti K, Sharma SK, Sundaram M, et al. Hippocampal changes in developing postnatal mice following intrauterine exposure to domoic acid. J Neurosci. 1993; 13:4486–4495. [PubMed: 8105041] Daneshian, M., Kamp, H., Hengstler, J., et al. Highlight report: Launch of a large integrated European in vitro toxicology project: EU-ToxRisk. Arch Toxicol. 2016. http://dx.doi.org/10.1007/ s00204-016-1698-7 Das KP, Freudenrich TM, Mundy WR. Assessment of PC12 cell differentiation and neurite growth: a comparison of morphological and neurochemical measures. Neurotoxicol Teratol. 2004; 26:397– 406. http://dx.doi.org/10.1016/j.ntt.2004.02.006. [PubMed: 15113601] Daston GP, Beyer BK, Carney EW, et al. Exposure-based validation list for developmental toxicity screening assays. Birth Defects Res B Dev Reprod Toxicol. 2014; 101:423–428. http://dx.doi.org/ 10.1002/bdrb.21132. [PubMed: 25475026] de Groot DM, Hartgring S, van de Horst L, et al. 2D and 3D assessment of neuropathology in rat brain after prenatal exposure to methylazoxymethanol, a model for developmental neurotoxicty. Reprod Toxicol. 2005; 20:417–432. http://dx.doi.org/10.1016/j.reprotox.2005.04.006. [PubMed: 15964739] Deoni SC, Mercure E, Blasi A, et al. Mapping infant brain myelination with magnetic resonance imaging. J Neurosci. 2011; 31:784–791. http://dx.doi.org/10.1523/JNEUROSCI.2106-10.2011. [PubMed: 21228187] Deskin R, Bursian SJ, Edens FW. Neurochemical alterations induced by manganese chloride in neonatal rats. Neurotoxicology. 1981; 2:65–73. [PubMed: 15622725] Deutsch SI, Burket JA, Katz E. Does subtle disturbance of neuronal migration contribute to schizophrenia and other neurodevelopmental disorders? Potential genetic mechanisms with possible treatment implications. Eur Neuropsychopharmacol. 2010; 20:281–287. http://dx.doi.org/ 10.1016/j.euroneuro.2010.02.005. [PubMed: 20207112] Dingemans MM, Ramakers GM, Gardoni F, et al. Neonatal exposure to brominated flame retardant BDE-47 reduces long-term potentiation and postsynaptic protein levels in mouse hippocampus. Environ Health Perspect. 2007; 115:865–870. http://dx.doi.org/10.1289/ehp.9860. [PubMed: 17589592] Doucette TA, Bernard PB, Yuill PC, et al. Low doses of non-NMDA glutamate receptor agonists alter neurobehavioural development in the rat. Neurotoxicol Teratol. 2003; 25:473–479. http:// dx.doi.org/10.1016/S0892-0362(03)00034-5. [PubMed: 12798964] Dreser N, Zimmer B, Dietz C, et al. Grouping of histone deacetylase inhibitors and other toxicants disturbing neural crest migration by transcriptional profiling. Neurotoxicology. 2015; 50:56–70. http://dx.doi.org/10.1016/j.neuro.2015.07.008. [PubMed: 26238599] Duarte-Salles T, von Stedingk H, Granum B, et al. Dietary acrylamide intake during pregnancy and fetal growth-results from the Norwegian mother and child cohort study (MoBa). Environ Health Perspect. 2013; 121:374–379. http://dx.doi.org/10.1289/ehp.1205396. [PubMed: 23204292]


Aschner et al.

Page 21


Dufault C, Poles G, Driscoll LL. Brief postnatal PBDE exposure alters learning and the cholinergic modulation of attention in rats. Toxicol Sci. 2005; 88:172–180. http://dx.doi.org/10.1093/toxsci/ kfi285. [PubMed: 16107551] Edwards SW, Tan YM, Villeneuve DL, et al. Adverse Outcome Pathways-Organizing Toxicological Information to Improve Decision Making. J Pharmacol Exp Ther. 2016; 356:170–181. http:// dx.doi.org/10.1124/jpet.115.228239. [PubMed: 26537250] Efremova L, Schildknecht S, Adam M, et al. Prevention of the degeneration of human dopaminergic neurons in an astrocyte co-culture system allowing endogenous drug metabolism. Br J Pharmacol. 2015; 172:4119–4132. http://dx.doi.org/10.1111/bph.13193. [PubMed: 25989025] Ekman L, Hansson E, Havu N, et al. Toxicological studies on omeprazole. Scand J Gastroenterol Suppl. 1985; 108:53–69. [PubMed: 3858976] Elsner J, Hodel B, Suter KE, et al. Detection limits of different approaches in behavioral teratology, and correlation of effects with neurochemical parameters. Neurotoxicol Teratol. 1988; 10:155– 167. http://dx.doi.org/10.1016/0892-0362(88)90080-3. [PubMed: 3398824] Eva R, Bouyoucef-Cherchalli D, Patel K, et al. IP3 3-kinase opposes NGF driven neurite outgrowth. PLoS One. 2012; 7:e32386. http://dx.doi.org/10.1371/journal.pone.0032386. [PubMed: 22384237] Falsig J, Latta M, Leist M. Defined inflammatory states in astrocyte cultures: correlation with susceptibility towards CD95-driven apoptosis. J Neurochem. 2004; 88:181–193. http://dx.doi.org/ 10.1111/j.1471-4159.2004.02144.x. [PubMed: 14675162] Fantel AG, Shepard TH, Newell-Morris LL, et al. Teratogenic effects of retinoic acid in pigtail monkeys (Macaca nemestrina). I. General features. Teratology. 1977; 15:65–71. http://dx.doi.org/ 10.1002/tera.1420150109. [PubMed: 402705] Forcelli PA, Janssen MJ, Vicini S, et al. Neonatal exposure to antiepileptic drugs disrupts striatal synaptic development. Ann Neurol. 2012; 72:363–372. http://dx.doi.org/10.1002/ana.23600. [PubMed: 22581672] Fredriksson A, Fredriksson M, Eriksson P. Neonatal exposure to paraquat or MPTP induces permanent changes in striatum dopamine and behavior in adult mice. Toxicol Appl Pharmacol. 1993; 122:258–264. http://dx.doi.org/10.1006/taap.1993.1194. [PubMed: 8212007] Fredriksson A, Ponten E, Gordh T, et al. Neonatal exposure to a combination of N-methyl-D-aspartate and gamma-aminobutyric acid type A receptor anesthetic agents potentiates apoptotic neurodegeneration and persistent behavioral deficits. Anesthesiology. 2007; 107:427–436. http:// dx.doi.org/10.1097/01.anes.0000278892.62305.9c. [PubMed: 17721245] Freeman JH Jr, Barone S Jr, Stanton ME. Cognitive and neuroanatomical effects of triethyltin in developing rats: role of age of exposure. Brain Res. 1994; 634:85–95. http://dx.doi.org/ 10.1016/0006-8993(94)90261-5. [PubMed: 8156395] Fritsche E, Cline JE, Nguyen NH, et al. Polychlorinated biphenyls disturb differentiation of normal human neural progenitor cells: clue for involvement of thyroid hormone receptors. Environ Health Perspect. 2005; 113:871–876. http://dx.doi.org/10.1289/ehp.7793. [PubMed: 16002375] Fryer SL, Mattson SN, Jernigan TL, et al. Caudate volume predicts neurocognitive performance in youth with heavy prenatal alcohol exposure. Alcohol Clin Exp Res. 2012; 36:1932–1941. http:// dx.doi.org/10.1111/j.1530-0277.2012.01811.x. [PubMed: 22551091] Gassmann K, Abel J, Bothe H, et al. Species-specific differential AhR expression protects human neural progenitor cells against developmental neurotoxicity of PAHs. Environ Health Perspect. 2010; 118:1571–1577. http://dx.doi.org/10.1289/ehp.0901545. [PubMed: 20570779] Gassmann K, Schreiber T, Dingemans MM, et al. BDE-47 and 6-OH-BDE-47 modulate calcium homeostasis in primary fetal human neural progenitor cells via ryanodine receptor-independent mechanisms. Arch Toxicol. 2014; 88:1537–1548. http://dx.doi.org/10.1007/s00204-014-1217-7. [PubMed: 24599297] Gatto CL, Broadie K. Genetic controls balancing excitatory and inhibitory synaptogenesis in neurodevelopmental disorder models. Front Synaptic Neurosci. 2010; 2:4. http://dx.doi.org/ 10.3389/fnsyn.2010.00004. [PubMed: 21423490] Gill SK, O’Brien L, Einarson TR, et al. The safety of proton pump inhibitors (PPIs) in pregnancy: a meta-analysis. Am J Gastroenterol. 2009; 104:1541–1545. quiz 1540, 1546. http://dx.doi.org/ 10.1038/ajg.2009.122. [PubMed: 19491869]


Aschner et al.

Page 22


Golub M, Kornetsky C. Effects of testing age and fostering experience on seizure susceptibility of rats treated prenatally with chlorpromazine. Dev Psychobiol. 1975; 8:519–524. http://dx.doi.org/ 10.1002/dev.420080608. [PubMed: 1233328] Grandjean P, Landrigan PJ. Developmental neurotoxicity of industrial chemicals. Lancet. 2006; 368:2167–2178. http://dx.doi.org/10.1016/S0140-6736(06)69665-7. [PubMed: 17174709] Grandjean P, Landrigan PJ. Neurobehavioural effects of developmental toxicity. Lancet Neurol. 2014; 13:330–338. http://dx.doi.org/10.1016/S1474-4422(13)70278-3. [PubMed: 24556010] Harrill JA, Freudenrich TM, Machacek DW, et al. Quantitative assessment of neurite outgrowth in human embryonic stem cell-derived hN2 cells using automated high-content image analysis. Neurotoxicology. 2010; 31:277–290. http://dx.doi.org/10.1016/j.neuro.2010.02.003. [PubMed: 20188755] Harrill JA, Freudenrich TM, Robinette BL, et al. Comparative sensitivity of human and rat neural cultures to chemical-induced inhibition of neurite outgrowth. Toxicol Appl Pharmacol. 2011a; 256:268–280. http://dx.doi.org/10.1016/j.taap.2011.02.013. [PubMed: 21354195] Harrill JA, Robinette BL, Mundy WR. Use of high content image analysis to detect chemical-induced changes in synaptogenesis in vitro. Toxicol In Vitro. 2011b; 25:368–387. http://dx.doi.org/ 10.1016/j.tiv.2010.10.011. [PubMed: 20969947] Hartung T, McBride M. Food for Thought … on mapping the human toxome. ALTEX. 2011; 28:83– 93. http://dx.doi.org/10.14573/altex.2011.2.083. [PubMed: 21625825] Hartung T, Hoffmann S, Stephens M. Mechanistic validation. ALTEX. 2013; 30:119–130. http:// dx.doi.org/10.14573/altex.2013.2.119. [PubMed: 23665802] Hass U, Lund SP, Hougaard KS, et al. Developmental neurotoxicity after toluene inhalation exposure in rats. Neurotoxicol Teratol. 1999; 21:349–357. http://dx.doi.org/10.1016/ S0892-0362(99)00013-6. [PubMed: 10440478] Hatta T, Ohmori H, Murakami T, et al. Neurotoxic effects of phenytoin on postnatal mouse brain development following neonatal administration. Neurotoxicol Teratol. 1999; 21:21–28. http:// dx.doi.org/10.1016/S0892-0362(98)00028-2. [PubMed: 10023798] Hirsch, C., Striegl, B., Mathes, S., et al. Multiparameter toxicity assessment of novel DOPO-derived organophosphorus flame retardants. Arch Toxicol. 2016. http://dx.doi.org/10.1007/ s00204-016-1680-4 Hoelting, L., Leist, M., Stoppini, L. Using Pluripotent Stem Cells and Their Progeny as an In Vitro Model to Assess (Developmental) Neurotoxicity. In: Pfannkuch, F., Suter-Dick, L., editors. Predictive Toxicology: From Vision to Reality. Weinheim, Germany: Wiley-VCH Verlag GmbH & Co. KGaA; 2015. Hogberg HT, Bal-Price AK. Domoic Acid-Induced Neurotoxicity Is Mainly Mediated by the AMPA/KA Receptor: Comparison between Immature and Mature Primary Cultures of Neurons and Glial Cells from Rat Cerebellum. J Toxicol. 2011; 2011:543512. http://dx.doi.org/ 10.1155/2011/543512. [PubMed: 22135676] Holson RR, Gazzara RA, Ferguson SA, et al. Behavioral effects of low-dose gestational day 11–13 retinoic acid exposure. Neurotoxicol Teratol. 1997; 19:355–362. http://dx.doi.org/10.1016/ S0892-0362(97)00041-X. [PubMed: 9380002] Hossain A, Hajman K, Charitidi K, et al. Prenatal dexamethasone impairs behavior and the activation of the BDNF exon IV promoter in the paraventricular nucleus in adult offspring. Endocrinology. 2008; 149:6356–6365. http://dx.doi.org/10.1210/en.2008-0388. [PubMed: 18755799] Hu Q, Fu H, Ren T, et al. Maternal low-level lead exposure reduces the expression of PSA-NCAM and the activity of sialyltransferase in the hippocampi of neonatal rat pups. Neurotoxicology. 2008; 29:675–681. http://dx.doi.org/10.1016/j.neuro.2008.04.002. [PubMed: 18499259] Infurna R, Weiss B. Neonatal behavioral toxicity in rats following prenatal exposure to methanol. Teratology. 1986; 33:259–265. http://dx.doi.org/10.1002/tera.1420330302. [PubMed: 3738821] Itahashi M, Abe H, Tanaka T, et al. Maternal exposure to hexachlorophene targets intermediate-stage progenitor cells of the hippocampal neurogenesis in rat offspring via dysfunction of cholinergic inputs by myelin vacuolation. Toxicology. 2015; 328:123–134. http://dx.doi.org/10.1016/j.tox. 2014.12.009. [PubMed: 25497112]


Aschner et al.

Page 23


Johansson N, Fredriksson A, Eriksson P. Neonatal exposure to perfluorooctane sulfonate (PFOS) and perfluorooctanoic acid (PFOA) causes neurobehavioural defects in adult mice. Neurotoxicology. 2008; 29:160–169. http://dx.doi.org/10.1016/j.neuro.2007.10.008. [PubMed: 18063051] Johnson FO, Chambers JE, Nail CA, et al. Developmental chlorpyrifos and methyl parathion exposure alters radial-arm maze performance in juvenile and adult rats. Toxicol Sci. 2009; 109:132–142. http://dx.doi.org/10.1093/toxsci/kfp053. [PubMed: 19293373] Johri A, Yadav S, Dhawan A, et al. Overexpression of cerebral and hepatic cytochrome P450s alters behavioral activity of rat offspring following prenatal exposure to lindane. Toxicol Appl Pharmacol. 2007; 225:278–292. http://dx.doi.org/10.1016/j.taap.2007.08.006. [PubMed: 17919674] Jones LB, Stanwood GD, Reinoso BS, et al. In utero cocaine-induced dysfunction of dopamine D1 receptor signaling and abnormal differentiation of cerebral cortical neurons. J Neurosci. 2000; 20:4606–4614. [PubMed: 10844030] Judson R, Houck K, Martin M, et al. In vitro and modelling approaches to risk assessment from the U.S. Environmental Protection Agency ToxCast programme. Basic Clin Pharmacol Toxicol. 2014; 115:69–76. http://dx.doi.org/10.1111/bcpt.12239. [PubMed: 24684691] Kabir ZD, Kennedy B, Katzman A, et al. Effects of prenatal cocaine exposure on social development in mice. Dev Neurosci. 2014; 36:338–346. http://dx.doi.org/10.1159/000360524. [PubMed: 24852757] Kadereit S, Zimmer B, van Thriel C, et al. Compound selection for in vitro modeling of developmental neurotoxicity. Front Biosci (Landmark Ed). 2012; 17:2442–2460. http://dx.doi.org/10.2741/4064. [PubMed: 22652791] Kamm JJ. Toxicology, carcinogenicity, and teratogenicity of some orally administered retinoids. J Am Acad Dermatol. 1982; 6:652–659. http://dx.doi.org/10.1016/S0190-9622(82)70054-4. [PubMed: 7040511] Kano Y, Nohno T, Hasegawa T, et al. Immunosuppressant FK506 induces neurite outgrowth in PC12 mutant cells with impaired NGF-promoted neuritogenesis via a novel MAP kinase signaling pathway. Neurochem Res. 2002; 27:1655–1661. http://dx.doi.org/10.1023/A:1021639128120. [PubMed: 12515319] Karlsson O, Jiang L, Ersson L, et al. Environmental neurotoxin interaction with proteins: Dosedependent increase of free and protein-associated BMAA (beta-N-methylamino-L-alanine) in neonatal rat brain. Sci Rep. 2015; 5:15570. http://dx.doi.org/10.1038/srep15570. [PubMed: 26498001] Kleensang A, Maertens A, Rosenberg M, et al. t4 workshop report: Pathways of Toxicity. ALTEX. 2014; 31:53–61. http://dx.doi.org/10.14573/altex.1309261. [PubMed: 24127042] Kochhar DM, Penner JD. Developmental effects of isotretinoin and 4-oxo-isotretinoin: the role of metabolism in teratogenicity. Teratology. 1987; 36:67–75. http://dx.doi.org/10.1002/tera. 1420360110. [PubMed: 3478842] Kristensson K, Eriksson H, Lundh B, et al. Effects of manganese chloride on the rat developing nervous system. Acta Pharmacol Toxicol (Copenh). 1986; 59:345–348. http://dx.doi.org/ 10.1111/j.1600-0773.1986.tb00182.x. [PubMed: 3811963] Krug AK, Balmer NV, Matt F, et al. Evaluation of a human neurite growth assay as specific screen for developmental neurotoxicants. Arch Toxicol. 2013a; 87:2215–2231. http://dx.doi.org/10.1007/ s00204-013-1072-y. [PubMed: 23670202] Krug AK, Kolde R, Gaspar JA, et al. Human embryonic stem cell-derived test systems for developmental neurotoxicity: a transcriptomics approach. Arch Toxicol. 2013b; 87:123–143. http://dx.doi.org/10.1007/s00204-012-0967-3. [PubMed: 23179753] Krug AK, Gutbier S, Zhao L, et al. Transcriptional and metabolic adaptation of human neurons to the mitochondrial toxicant MPP(+). Cell Death Dis. 2014; 5:e1222. http://dx.doi.org/10.1038/cddis. 2014.166. [PubMed: 24810058] Kuegler PB, Zimmer B, Waldmann T, et al. Markers of murine embryonic and neural stem cells, neurons and astrocytes: reference points for developmental neurotoxicity testing. ALTEX. 2010; 27:17–42. [PubMed: 20390237]


Aschner et al.

Page 24


Kuegler PB, Baumann BA, Zimmer B, et al. GFAP-independent inflammatory competence and trophic functions of astrocytes generated from murine embryonic stem cells. Glia. 2012; 60:218–228. http://dx.doi.org/10.1002/glia.21257. [PubMed: 22072312] Langeveld WT, Meijer M, Westerink RH. Differential effects of 20 non-dioxin-like PCBs on basal and depolarization-evoked intracellular calcium levels in PC12 cells. Toxicol Sci. 2012; 126:487– 496. http://dx.doi.org/10.1093/toxsci/kfr346. [PubMed: 22218490] Lasky DI, Zagon IS, McLaughlin PJ. Effect of maternally administered heroin on the motor activity of rat offspring. Pharmacol Biochem Behav. 1977; 7:281–284. http://dx.doi.org/ 10.1016/0091-3057(77)90147-2. [PubMed: 928485] Lein P, Silbergeld E, Locke P, et al. In vitro and other alternative approaches to developmental neurotoxicity testing (DNT). Environ Toxicol Pharmacol. 2005; 19:735–744. http://dx.doi.org/ 10.1016/j.etap.2004.12.035. [PubMed: 21783550] Lein PJ, Yang D, Bachstetter AD, et al. Ontogenetic alterations in molecular and structural correlates of dendritic growth after developmental exposure to polychlorinated biphenyls. Environ Health Perspect. 2007; 115:556–563. http://dx.doi.org/10.1289/ehp.9773. [PubMed: 17450224] Leist M, Efremova L, Karreman C. Food for thought … considerations and guidelines for basic test method descriptions in toxicology. ALTEX. 2010; 27:309–317. [PubMed: 21240472] Leist M, Hasiwa N, Daneshian M, et al. Validation and quality control of replacement alternatives – current status and future challenges. Toxicology Research. 2012; 1:8–22. http://dx.doi.org/ 10.1039/C2TX20011B. Leist M, Hasiwa N, Rovida C, et al. Consensus report on the future of animal-free systemic toxicity testing. ALTEX. 2014; 31:341–356. http://dx.doi.org/http://dx.doi.org/10.14573/altex.1406091. [PubMed: 25061899] LeSage MG, Gustaf E, Dufek MB, et al. Effects of maternal intravenous nicotine administration on locomotor behavior in pre-weanling rats. Pharmacol Biochem Behav. 2006; 85:575–583. http:// dx.doi.org/10.1016/j.pbb.2006.10.012. [PubMed: 17141848] Levin ED, Briggs SJ, Christopher NC, et al. Prenatal nicotine exposure and cognitive performance in rats. Neurotoxicol Teratol. 1993; 15:251–260. http://dx.doi.org/10.1016/0892-0362(93)90006-A. [PubMed: 8413079] Levin ED, Pizarro K, Pang WG, et al. Persisting behavioral consequences of prenatal domoic acid exposure in rats. Neurotoxicol Teratol. 2005; 27:719–725. http://dx.doi.org/10.1016/j.ntt. 2005.06.017. [PubMed: 16054336] Li AA, Lowe KA, McIntosh LJ, et al. Evaluation of epidemiology and animal data for risk assessment: chlorpyrifos developmental neurobehavioral outcomes. J Toxicol Environ Health B Crit Rev. 2012; 15:109–184. http://dx.doi.org/10.1080/10937404.2012.645142. [PubMed: 22401178] Liew Z, Ritz B, Olsen J. Characteristics of acetaminophen users compared with nonusers during pregnancy, behavioral problems, and hyperkinetic disorders–reply. JAMA Pediatr. 2014; 168:865–866. http://dx.doi.org/10.1001/jamapediatrics.2014.983. Liu J, Zheng X, Yin F, et al. Neurotrophic property of geniposide for inducing the neuronal differentiation of PC12 cells. Int J Dev Neurosci. 2006; 24:419–424. http://dx.doi.org/10.1016/ j.ijdevneu.2006.08.009. [PubMed: 17045447] Lown BA, Morganti JB, D’Agostino R, et al. Effects on the postnatal development of the mouse of preconception, postconception and/or suckling exposure to manganese via maternal inhalation exposure to MnO2 dust. Neurotoxicology. 1984; 5:119–129. [PubMed: 6538947] Lu R, Liu X, Long H, et al. Effects of prenatal cocaine and heroin exposure on neuronal dendrite morphogenesis and spatial recognition memory in mice. Neurosci Lett. 2012; 522:128–133. http://dx.doi.org/10.1016/j.neulet.2012.06.023. [PubMed: 22732446] Lucchi L, Covelli V, Petkov VV, et al. Effects of ethanol, given during pregnancy, on the offspring dopaminergic system. Pharmacol Biochem Behav. 1983; 19:567–570. http://dx.doi.org/ 10.1016/0091-3057(83)90328-3. [PubMed: 6316369] Mactutus CF, Harrod SB, Hord LL, et al. Prenatal IV Cocaine: Alterations in Auditory Information Processing. Front Psychiatry. 2011; 2:38. http://dx.doi.org/10.3389/fpsyt.2011.00038. [PubMed: 21747770]


Aschner et al.

Page 25


Mahony C, Erskine L, Niven J, et al. Pomalidomide is nonteratogenic in chicken and zebrafish embryos and nonneurotoxic in vitro. Proc Natl Acad Sci U S A. 2013; 110:12703–12708. http:// dx.doi.org/10.1073/pnas.1307684110. [PubMed: 23858438] Mandell JW, Banker GA. Selective blockade of axonogenesis in cultured hippocampal neurons by the tyrosine phosphatase inhibitor orthovanadate. J Neurobiol. 1998; 35:17–28. http://dx.doi.org/ 10.1002/(SICI)1097-4695(199804)35:13.0.CO;2-E. [PubMed: 9552163] Martinez-Finley EJ, Ali AM, Allan AM. Learning deficits in C57BL/6J mice following perinatal arsenic exposure: consequence of lower corticosterone receptor levels? Pharmacol Biochem Behav. 2009; 94:271–277. http://dx.doi.org/10.1016/j.pbb.2009.09.006. [PubMed: 19751756] Maurissen JP, Hoberman AM, Garman RH, et al. Lack of selective developmental neurotoxicity in rat pups from dams treated by gavage with chlorpyrifos. Toxicol Sci. 2000; 57:250–263. http:// dx.doi.org/10.1093/toxsci/57.2.250. [PubMed: 11006355] McCartney MA, Scinto PL, Wang SS, et al. Developmental effects of phenytoin may differ depending on sex of offspring. Neurotoxicol Teratol. 1999; 21:119–128. http://dx.doi.org/10.1016/ S0892-0362(98)00047-6. [PubMed: 10192272] McElhatton PR, Bateman DN, Evans C, et al. Congenital anomalies after prenatal ecstasy exposure. Lancet. 1999; 354:1441–1442. http://dx.doi.org/10.1016/S0140-6736(99)02423-X. [PubMed: 10543673] Meijer M, Dingemans MM, van den Berg M, et al. Inhibition of voltage-gated calcium channels as common mode of action for (mixtures of) distinct classes of insecticides. Toxicol Sci. 2014a; 141:103–111. http://dx.doi.org/10.1093/toxsci/kfu110. [PubMed: 24913802] Meijer M, Hamers T, Westerink RH. Acute disturbance of calcium homeostasis in PC12 cells as a novel mechanism of action for (sub)micromolar concentrations of organophosphate insecticides. Neurotoxicology. 2014b; 43:110–116. http://dx.doi.org/10.1016/j.neuro.2014.01.008. [PubMed: 24495583] Miller GW. Improving reproducibility in toxicology. Toxicol Sci. 2014; 139:1–3. http://dx.doi.org/ 10.1093/toxsci/kfu050. [PubMed: 24747876] Moors M, Cline JE, Abel J, et al. ERK-dependent and -independent pathways trigger human neural progenitor cell migration. Toxicol Appl Pharmacol. 2007; 221:57–67. http://dx.doi.org/10.1016/ j.taap.2007.02.018. [PubMed: 17445854] Moors M, Rockel TD, Abel J, et al. Human neurospheres as three-dimensional cellular systems for developmental neurotoxicity testing. Environ Health Perspect. 2009; 117:1131–1138. http:// dx.doi.org/10.1289/ehp.0800207. [PubMed: 19654924] Moors M, Vudattu NK, Abel J, et al. Interleukin-7 (IL-7) and IL-7 splice variants affect differentiation of human neural progenitor cells. Genes Immun. 2010; 11:11–20. http://dx.doi.org/10.1038/gene. 2009.77. [PubMed: 19847194] Mount DL, Feeney P, Fabricatore AN, et al. Constructing common cohorts from trials with overlapping eligibility criteria: implications for comparing effect sizes between trials. Clin Trials. 2009; 6:416–429. http://dx.doi.org/10.1177/1740774509344440. [PubMed: 19737845] Mundy WR, Radio NM, Freudenrich TM. Neuronal models for evaluation of proliferation in vitro using high content screening. Toxicology. 2010; 270:121–130. http://dx.doi.org/10.1016/j.tox. 2010.02.004. [PubMed: 20149836] Mundy WR, Padilla S, Breier JM, et al. Expanding the test set: Chemicals with potential to disrupt mammalian brain development. Neurotoxicol Teratol. 2015; 52:25–35. http://dx.doi.org/10.1016/ j.ntt.2015.10.001. [PubMed: 26476195] Nau H. Species differences in pharmacokinetics and drug teratogenesis. Environ Health Perspect. 1986; 70:113–129. http://dx.doi.org/10.1289/ehp.8670113. [PubMed: 3104022] Nau H. Teratogenicity of isotretinoin revisited: species variation and the role of all-trans-retinoic acid. J Am Acad Dermatol. 2001; 45:S183–187. http://dx.doi.org/10.1067/mjd.2001.113720. [PubMed: 11606951] Nicolai J, Vles JS, Aldenkamp AP. Neurodevelopmental delay in children exposed to antiepileptic drugs in utero: a critical review directed at structural study-bias. J Neurol Sci. 2008; 271:1–14. http://dx.doi.org/10.1016/j.jns.2008.03.004. [PubMed: 18479711]


Aschner et al.

Page 26


Niebyl, J., Simpson, J. Teratology and Drugs in Pregnancy. Vol Glob libr women’s med. 2008. http:// dx.doi.org/10.3843/GLOWM.10096 Nolen GA. The effects of prenatal retinoic acid on the viability and behavior of the offspring. Neurobehav Toxicol Teratol. 1986; 8:643–654. [PubMed: 3808180] O’Callaghan JP, Miller DB, Reiter LW. Acute postnatal exposure to triethyltin in the rat: effects on specific protein composition of subcellular fractions from developing and adult brain. J Pharmacol Exp Ther. 1983; 224:466–472. [PubMed: 6822967] O’Connor TG, Bergman K, Sarkar P, et al. Prenatal cortisol exposure predicts infant cortisol response to acute stress. Dev Psychobiol. 2013; 55:145–155. http://dx.doi.org/10.1002/dev.21007. [PubMed: 22315044] Ochi N, Naoi M, Mogi M, et al. Effects of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) administration in prenatal stage on the dopamine system in the postnatal mouse brain. Life Sci. 1991; 48:217–223. http://dx.doi.org/10.1016/0024-3205(91)90348-F. [PubMed: 1671518] Onishchenko N, Fischer C, Wan Ibrahim WN, et al. Prenatal exposure to PFOS or PFOA alters motor function in mice in a sex-related manner. Neurotox Res. 2011; 19:452–461. http://dx.doi.org/ 10.1007/s12640-010-9200-4. [PubMed: 20512442] Oshiro WM, Beasley TE, McDaniel KL, et al. Selective cognitive deficits in adult rats after prenatal exposure to inhaled ethanol. Neurotoxicol Teratol. 2014; 45:44–58. http://dx.doi.org/10.1016/ j.ntt.2014.07.001. [PubMed: 25020118] Owens MY, Wallace KL, Mamoon N, et al. Absence of neurotoxicity with medicinal grade terbutaline in the rat model. Reprod Toxicol. 2011; 31:447–453. http://dx.doi.org/10.1016/j.reprotox. 2011.01.001. [PubMed: 21262341] Pallocca, G., Grinberg, M., Henry, M., et al. Identification of transcriptome signatures and biomarkers specific for potential developmental toxicants inhibiting human neural crest cell migration. Arch Toxicol. 2015. http://dx.doi.org/10.1007/s00204-015-1658-7 Pallocca G, Grinberg M, Henry M, et al. Identification of transcriptome signatures and biomarkers specific for potential developmental toxicants inhibiting human neural crest cell migration. Arch Toxicol. 2016; 90:159–180. http://dx.doi.org/10.1007/s00204-015-1658-7. [PubMed: 26705709] Parran DK, Mundy WR, Barone S Jr. Effects of methylmercury and mercuric chloride on differentiation and cell viability in PC12 cells. Toxicol Sci. 2001; 59:278–290. http://dx.doi.org/ 10.1093/toxsci/59.2.278. [PubMed: 11158721] Paule MG, Li M, Allen RR, et al. Ketamine anesthesia during the first week of life can cause longlasting cognitive deficits in rhesus monkeys. Neurotoxicol Teratol. 2011; 33:220–230. http:// dx.doi.org/10.1016/j.ntt.2011.01.001. [PubMed: 21241795] Pedersen M, von Stedingk H, Botsivali M, et al. Birth weight, head circumference, and prenatal exposure to acrylamide from maternal diet: the European prospective mother-child study (NewGeneris). Environ Health Perspect. 2012; 120:1739–1745. http://dx.doi.org/10.1289/ehp. 1205327. [PubMed: 23092936] Pei, Y., Peng, J., Behl, M., et al. Comparative neurotoxicity screening in human iPSC-derived neural stem cells, neurons and astrocytes. Brain Res. 2015. http://dx.doi.org/10.1016/j.brainres. 2015.07.048 Penschuck S, Flagstad P, Didriksen M, et al. Decrease in parvalbumin-expressing neurons in the hippocampus and increased phencyclidine-induced locomotor activity in the rat methylazoxymethanol (MAM) model of schizophrenia. Eur J Neurosci. 2006; 23:279–284. http:// dx.doi.org/10.1111/j.1460-9568.2005.04536.x. [PubMed: 16420437] Perkins EJ, Antczak P, Burgoon L, et al. Adverse Outcome Pathways for Regulatory Applications: Examination of Four Case Studies With Different Degrees of Completeness and Scientific Confidence. Toxicol Sci. 2015; 148:14–25. http://dx.doi.org/10.1093/toxsci/kfv181. [PubMed: 26500288] Petit TL, LeBoutillier JC, Brooks WJ. Altered sensitivity to NMDA following developmental lead exposure in rats. Physiol Behav. 1992; 52:687–693. http://dx.doi.org/ 10.1016/0031-9384(92)90398-L. [PubMed: 1409940] Poland CA, Miller MR, Duffin R, et al. The elephant in the room: reproducibility in toxicology. Part Fibre Toxicol. 2014; 11:42. http://dx.doi.org/10.1186/s12989-014-0042-8. [PubMed: 25149182]


Aschner et al.

Page 27


Radio NM, Breier JM, Shafer TJ, et al. Assessment of chemical effects on neurite outgrowth in PC12 cells using high content screening. Toxicol Sci. 2008; 105:106–118. http://dx.doi.org/10.1093/ toxsci/kfn114. [PubMed: 18539913] Radio NM, Freudenrich TM, Robinette BL, et al. Comparison of PC12 and cerebellar granule cell cultures for evaluating neurite outgrowth using high content analysis. Neurotoxicol Teratol. 2010; 32:25–35. http://dx.doi.org/10.1016/j.ntt.2009.06.003. [PubMed: 19559085] Radonjic M, Cappaert NL, de Vries EF, et al. Delay and impairment in brain development and function in rat offspring after maternal exposure to methylmercury. Toxicol Sci. 2013; 133:112–124. http://dx.doi.org/10.1093/toxsci/kft024. [PubMed: 23457123] Ramirez T, Daneshian M, Kamp H, et al. Metabolomics in toxicology and preclinical research. ALTEX. 2013; 30:209–225. http://dx.doi.org/10.14573/altex.2013.2.209. [PubMed: 23665807] Reddy GR, Devi BC, Chetty CS. Developmental lead neurotoxicity: alterations in brain cholinergic system. Neurotoxicology. 2007; 28:402–407. http://dx.doi.org/10.1016/j.neuro.2006.03.018. [PubMed: 16678265] Reel JR, Lawton AD, Lamb JCt. Reproductive toxicity evaluation of acetaminophen in Swiss CD-1 mice using a continuous breeding protocol. Fundam Appl Toxicol. 1992; 18:233–239. http:// dx.doi.org/10.1016/0272-0590(92)90051-I. [PubMed: 1601223] Rempel E, Hoelting L, Waldmann T, et al. A transcriptome-based classifier to identify developmental toxicants by stem cell testing: design, validation and optimization for histone deacetylase inhibitors. Arch Toxicol. 2015; 89:1599–1618. http://dx.doi.org/10.1007/s00204-015-1573-y. [PubMed: 26272509] Rivera S, Sanfeliu C, Rodriguez-Farre E. Behavioral changes induced in developing rats by an early postnatal exposure to lindane. Neurotoxicol Teratol. 1990; 12:591–595. http://dx.doi.org/ 10.1016/0892-0362(90)90067-M. [PubMed: 1701515] Robertson RT, Majka JA, Peter CP, et al. Effects of prenatal exposure to chlorpromazine on postnatal development and behavior of rats. Toxicol Appl Pharmacol. 1980; 53:541–549. http://dx.doi.org/ 10.1016/0041-008X(80)90367-1. [PubMed: 7385249] Robinette BL, Harrill JA, Mundy WR, et al. In vitro assessment of developmental neurotoxicity: use of microelectrode arrays to measure functional changes in neuronal network ontogeny. Front Neuroeng. 2011; 4:1. http://dx.doi.org/10.3389/fneng.2011.00001. [PubMed: 21270946] Rodriguez VM, Carrizales L, Mendoza MS, et al. Effects of sodium arsenite exposure on development and behavior in the rat. Neurotoxicol Teratol. 2002; 24:743–750. http://dx.doi.org/10.1016/ S0892-0362(02)00313-6. [PubMed: 12460656] Rosengarten H, Quartermain D. Effect of prenatal administration of haloperidol, risperidone, quetiapine and olanzapine on spatial learning and retention in adult rats. Pharmacol Biochem Behav. 2002; 72:575–579. http://dx.doi.org/10.1016/S0091-3057(02)00727-X. [PubMed: 12175454] Rovida C, Longo F, Rabbit RR. How are reproductive toxicity and developmental toxicity addressed in REACH dossiers? ALTEX. 2011; 28:273–294. http://dx.doi.org/10.14573/altex.2011.4.273. [PubMed: 22130481] Rovida C, Alepee N, Api AM, et al. Integrated Testing Strategies (ITS) for safety assessment. ALTEX. 2015; 32:25–40. http://dx.doi.org/http://dx.doi.org/10.14573/altex.1411011. [PubMed: 25413849] Ryan KR, Sirenko O, Parham F, et al. Neurite outgrowth in human induced pluripotent stem cellderived neurons as a high-throughput screen for developmental neurotoxicity or neurotoxicity. Neurotoxicology. 2016; 53:271–281. http://dx.doi.org/10.1016/j.neuro.2016.02.003. [PubMed: 26854185] Sable HJ, Powers BE, Wang VC, et al. Alterations in DRH and DRL performance in rats developmentally exposed to an environmental PCB mixture. Neurotoxicol Teratol. 2006; 28:548– 556. http://dx.doi.org/10.1016/j.ntt.2006.06.005. [PubMed: 16930942] Sakamoto M, Kakita A, Wakabayashi K, et al. Evaluation of changes in methylmercury accumulation in the developing rat brain and its effects: a study with consecutive and moderate dose exposure throughout gestation and lactation periods. Brain Res. 2002; 949:51–59. http://dx.doi.org/ 10.1016/S0006-8993(02)02964-5. [PubMed: 12213299]


Aschner et al.

Page 28


Sauer JM, Hartung T, Leist M, et al. Systems Toxicology: The Future of Risk Assessment. Int J Toxicol. 2015; 34:346–348. http://dx.doi.org/10.1177/1091581815576551. [PubMed: 25804424] Schildknecht S, Pape R, Meiser J, et al. Preferential Extracellular Generation of the Active Parkinsonian Toxin MPP(+) by Transporter-Independent Export of the Intermediate MPDP(.). Antioxid Redox Signal. 2015; 23:1001–1016. http://dx.doi.org/10.1089/ars.2015.6297. [PubMed: 26413876] Schneider T, Przewlocki R. Behavioral alterations in rats prenatally exposed to valproic acid: animal model of autism. Neuropsychopharmacology. 2005; 30:80–89. http://dx.doi.org/10.1038/sj.npp. 1300518. [PubMed: 15238991] Schreiber T, Gassmann K, Gotz C, et al. Polybrominated diphenyl ethers induce developmental neurotoxicity in a human in vitro model: evidence for endocrine disruption. Environ Health Perspect. 2010; 118:572–578. http://dx.doi.org/10.1289/ehp.0901435. [PubMed: 20368126] Semrud-Clikeman, M., Ellison, T. Child Neuropsychology: Assessment and Interventions for Neurodevelopmental Disorders. Vol. 2. Springer; US: 2009. http://dx.doi.org/ 10.1007/978-0-387-88963-4 Shafer TJ, Crofton KM. Comments on: ‘Perinatal toxicity of cyfluthrin in mice: developmental and behavioral effects’ by Soni and colleagues. Hum Exp Toxicol. 2011; 30:1112–3. http:// dx.doi.org/10.1177/0960327111411500. [PubMed: 21771863] Shafer TJ, Meyer DA, Crofton KM. Developmental neurotoxicity of pyrethroid insecticides: critical review and future research needs. Environ Health Perspect. 2005; 113:123–36. http://dx.doi.org/ 10.1289/ehp.7254. [PubMed: 15687048] Sheets LP, Li AA, Minnema DJ, et al. A critical review of neonicotinoid insecticides for developmental neurotoxicity. Crit Rev Toxicol. 2016; 46:153–190. http://dx.doi.org/ 10.3109/10408444.2015.1090948. [PubMed: 26513508] Shuman RM, Leech RW, Alvord EC Jr. Neurotoxicity of hexachlorophene in the human: I. A clinicopathologic study of 248 children. Pediatrics. 1974; 54:689–695. [PubMed: 4431666] Singer LT, Moore DG, Min MO, et al. One-year outcomes of prenatal exposure to MDMA and other recreational drugs. Pediatrics. 2012; 130:407–413. http://dx.doi.org/10.1542/peds.2012-0666. [PubMed: 22908109] Smirnova L, Hogberg HT, Leist M, et al. Developmental neurotoxicity – challenges in the 21st century and in vitro opportunities. ALTEX. 2014; 31:129–156. http://dx.doi.org/http://dx.doi.org/ 10.14573/altex.1403271. [PubMed: 24687333] Smirnova, L., Harris, G., Delp, J., et al. A LUHMES 3D dopaminergic neuronal model for neurotoxicity testing allowing long-term exposure and cellular resilience analysis. Arch Toxicol. 2015. http://dx.doi.org/10.1007/s00204-015-1637-z Sobotka TJ, Brodie RE, Cook MP. Behavioral and neuroendocrine effects in rats of postnatal exposure to low dietary levels of maneb. Dev Psychobiol. 1972; 5:137–148. http://dx.doi.org/10.1002/dev. 420050207. [PubMed: 4671406] Stern S, Cox C, Preston R, et al. Perinatal methanol exposure in the rat. II. Behavioral effects in neonates and adults. Fundam Appl Toxicol. 1997; 36:163–176. http://dx.doi.org/10.1006/faat. 1997.2288. [PubMed: 9143486] Stiegler NV, Krug AK, Matt F, et al. Assessment of chemical-induced impairment of human neurite outgrowth by multiparametric live cell imaging in high-density cultures. Toxicol Sci. 2011; 121:73–87. http://dx.doi.org/10.1093/toxsci/kfr034. [PubMed: 21342877] Sullivan-Jones P, Ali SF, Gough B, et al. Postnatal methylazoxymethanol: sensitive periods and regional selectivity of effects. Neurotoxicol Teratol. 1994; 16:631–637. http://dx.doi.org/ 10.1016/0892-0362(94)90041-8. [PubMed: 7862061] Tegenge MA, Rockel TD, Fritsche E, et al. Nitric oxide stimulates human neural progenitor cell migration via cGMP-mediated signal transduction. Cell Mol Life Sci. 2011; 68:2089–2099. http://dx.doi.org/10.1007/s00018-010-0554-9. [PubMed: 20957508] Thiruchelvam M, Richfield EK, Goodman BM, et al. Developmental exposure to the pesticides paraquat and maneb and the Parkinson’s disease phenotype. Neurotoxicology. 2002; 23:621–633. http://dx.doi.org/10.1016/S0161-813X(02)00092-X. [PubMed: 12428734]


Aschner et al.

Page 29


Thomas JH. Nematodes are smarter than you think. Neuron. 2001; 30:7–8. http://dx.doi.org/10.1016/ S0896-6273(01)00256-2. [PubMed: 11343638] Thompson VB, Koprich JB, Chen EY, et al. Prenatal exposure to MDMA alters noradrenergic neurodevelopment in the rat. Neurotoxicol Teratol. 2012; 34:206–213. http://dx.doi.org/10.1016/ j.ntt.2011.09.005. [PubMed: 21978916] Thor S. The genetics of brain development: conserved programs in flies and mice. Neuron. 1995; 15:975–977. http://dx.doi.org/10.1016/0896-6273(95)90084-5. [PubMed: 7576663] Tolins M, Ruchirawat M, Landrigan P. The developmental neurotoxicity of arsenic: cognitive and behavioral consequences of early life exposure. Ann Glob Health. 2014; 80:303–314. http:// dx.doi.org/10.1016/j.aogh.2014.09.005. [PubMed: 25459332] Tollefsen KE, Scholz S, Cronin MT, et al. Applying Adverse Outcome Pathways (AOPs) to support Integrated Approaches to Testing and Assessment (IATA). Regul Toxicol Pharmacol. 2014; 70:629–640. http://dx.doi.org/10.1016/j.yrtph.2014.09.009. [PubMed: 25261300] Tropepe V, Sive HL. Can zebrafish be used as a model to study the neurodevelopmental causes of autism? Genes Brain Behav. 2003; 2:268–281. http://dx.doi.org/10.1034/j.1601-183X. 2003.00038.x. [PubMed: 14606692] Ulsamer AG, Yoder PD, Kimbrough RD, et al. Effects of hexachlorophene on developing rats: toxicity, tissue concentrations and biochemistry. Food Cosmet Toxicol. 1975; 13:69–80. http://dx.doi.org/ 10.1016/0015-6264(75)90084-X. [PubMed: 1123204] van Thriel C, Westerink RH, Beste C, et al. Translating neurobehavioural endpoints of developmental neurotoxicity tests into in vitro assays and readouts. Neurotoxicology. 2012; 33:911–924. http:// dx.doi.org/10.1016/j.neuro.2011.10.002. [PubMed: 22008243] van Vliet E, Daneshian M, Beilmann M, et al. Current approaches and future role of high content imaging in safety sciences and drug discovery. ALTEX. 2014; 31:479–493. http://dx.doi.org/ http://dx.doi.org/10.14573/altex.1405271. [PubMed: 25027442] Viberg H, Fredriksson A, Jakobsson E, et al. Neurobehavioral derangements in adult mice receiving decabrominated diphenyl ether (PBDE 209) during a defined period of neonatal brain development. Toxicol Sci. 2003; 76:112–120. http://dx.doi.org/10.1093/toxsci/kfg210. [PubMed: 12915714] Viberg H, Eriksson P, Gordh T, et al. Paracetamol (acetaminophen) administration during neonatal brain development affects cognitive function and alters its analgesic and anxiolytic response in adult male mice. Toxicol Sci. 2014; 138:139–147. http://dx.doi.org/10.1093/toxsci/kft329. [PubMed: 24361869] Vorhees CV. Behavioral teratogenicity of valproic acid: selective effects on behavior after prenatal exposure to rats. Psychopharmacology (Berl). 1987; 92:173–179. http://dx.doi.org/10.1007/ BF00177911. [PubMed: 3110838] Vorhees CV. Developmental neurotoxicity induced by therapeutic and illicit drugs. Environ Health Perspect. 1994; 102(Suppl 2):145–153. http://dx.doi.org/10.1289/ehp.94102145. Waldmann T, Rempel E, Balmer NV, et al. Design principles of concentration-dependent transcriptome deviations in drug-exposed differentiating stem cells. Chem Res Toxicol. 2014; 27:408–420. http://dx.doi.org/10.1021/tx400402j. [PubMed: 24383497] Wang Y, Han TZ. Prenatal exposure to heroin in mice elicits memory deficits that can be attributed to neuronal apoptosis. Neuroscience. 2009; 160:330–338. http://dx.doi.org/10.1016/j.neuroscience. 2009.02.058. [PubMed: 19272431] Watanabe KH, Andersen ME, Basu N, et al. Defining and modeling known adverse outcome pathways: Domoic acid and neuronal signaling as a case study. Environ Toxicol Chem. 2011; 30:9–21. http://dx.doi.org/10.1002/etc.373. [PubMed: 20963854] Watanabe T, Matsuhashi K, Takayama S. Study on the postnatal neuro-behavioral development in rats treated prenatally with drugs acting on the autonomic nervous systems. Nihon Yakurigaku Zasshi. 1985; 85:79–90. http://dx.doi.org/10.1254/fpj.85.79. [PubMed: 3988167] Weisenburger WP, Minck DR, Acuff KD, et al. Dose-response effects of prenatal phenytoin exposure in rats: effects on early locomotion, maze learning, and memory as a function of phenytoininduced circling behavior. Neurotoxicol Teratol. 1990; 12:145–152. http://dx.doi.org/ 10.1016/0892-0362(90)90127-X. [PubMed: 2333067]


Aschner et al.

Page 30


Westerink RH. Do we really want to REACH out to in vitro? Neurotoxicology. 2013; 39:169–172. http://dx.doi.org/10.1016/j.neuro.2013.10.001. [PubMed: 24125872] Wolansky MJ, Soiza-Reilly M, Fossati M, et al. Postnatal haloperidol eliminates the deficit in circling behavior produced by prenatal exposure to the same drug. Neurotoxicol Teratol. 2004; 26:561– 569. http://dx.doi.org/10.1016/j.ntt.2004.04.006. [PubMed: 15203178] Yanai J, Avraham Y, Levy S, et al. Alterations in septohippocampal cholinergic innervations and related behaviors after early exposure to heroin and phencyclidine. Brain Res Dev Brain Res. 1992; 69:207–214. http://dx.doi.org/10.1016/0165-3806(92)90161-O. [PubMed: 1424097] Yang D, Howard A, Bruun D, et al. Chlorpyrifos and chlorpyrifos-oxon inhibit axonal growth by interfering with the morphogenic activity of acetylcholinesterase. Toxicol Appl Pharmacol. 2008; 228:32–41. http://dx.doi.org/10.1016/j.taap.2007.11.005. [PubMed: 18076960] Yang D, Kim KH, Phimister A, et al. Developmental exposure to polychlorinated biphenyls interferes with experience-dependent dendritic plasticity and ryanodine receptor expression in weanling rats. Environ Health Perspect. 2009; 117:426–435. http://dx.doi.org/10.1289/ehp.11771. [PubMed: 19337518] Yang D, Kania-Korwel I, Ghogha A, et al. PCB 136 atropselectively alters morphometric and functional parameters of neuronal connectivity in cultured rat hippocampal neurons via ryanodine receptor-dependent mechanisms. Toxicol Sci. 2014; 138:379–392. http://dx.doi.org/10.1093/ toxsci/kft334. [PubMed: 24385416] Yolton K, Cornelius M, Ornoy A, et al. Exposure to neurotoxicants and the development of attention deficit hyperactivity disorder and its related behaviors in childhood. Neurotoxicol Teratol. 2014; 44:30–45. http://dx.doi.org/10.1016/j.ntt.2014.05.003. [PubMed: 24846602] Zeng HC, Li YY, Zhang L, et al. Prenatal exposure to perfluorooctanesulfonate in rat resulted in longlasting changes of expression of synapsins and synaptophysin. Synapse. 2011; 65:225–233. http://dx.doi.org/10.1002/syn.20840. [PubMed: 20687110] Zerrate MC, Pletnikov M, Connors SL, et al. Neuroinflammation and behavioral abnormalities after neonatal terbutaline treatment in rats: implications for autism. J Pharmacol Exp Ther. 2007; 322:16–22. http://dx.doi.org/10.1124/jpet.107.121483. [PubMed: 17400887] Zhao T, Li Y, Wei W, et al. Ketamine administered to pregnant rats in the second trimester causes longlasting behavioral disorders in offspring. Neurobiol Dis. 2014; 68:145–155. http://dx.doi.org/ 10.1016/j.nbd.2014.02.009. [PubMed: 24780497] Zimmer B, Kuegler PB, Baudis B, et al. Coordinated waves of gene expression during neuronal differentiation of embryonic stem cells as basis for novel approaches to developmental neurotoxicity testing. Cell Death Differ. 2011a; 18:383–395. http://dx.doi.org/10.1038/cdd. 2010.109. [PubMed: 20865013] Zimmer B, Schildknecht S, Kuegler PB, et al. Sensitivity of dopaminergic neuron differentiation from stem cells to chronic low-dose methylmercury exposure. Toxicol Sci. 2011b; 121:357–367. http:// dx.doi.org/10.1093/toxsci/kfr054. [PubMed: 21385734] Zimmer B, Lee G, Balmer NV, et al. Evaluation of developmental toxicants and signaling pathways in a functional test based on the migration of human neural crest cells. Environ Health Perspect. 2012; 120:1116–1122. http://dx.doi.org/10.1289/ehp.1104489. [PubMed: 22571897] Zimmer B, Pallocca G, Dreser N, et al. Profiling of drugs and environmental chemicals for functional impairment of neural crest migration in a novel stem cell-based test battery. Arch Toxicol. 2014; 88:1109–1126. http://dx.doi.org/10.1007/s00204-014-1231-9. [PubMed: 24691702]

Author Manuscript ALTEX. Author manuscript; available in PMC 2017 April 01.

Aschner et al.

Page 31

Author Manuscript

Box 1 Glossary for assay definition and setup Test methods vs test systems Test system

Author Manuscript

Cellular (or biochemical) system used for a test method (e.g. “proliferating hESC”, or “neuronally-differentiating PC-12 cells”, or “organotypic brain slices”). The term is often used interchangeably with “in vitro system”, or sometimes also termed “biological model”. The test system is only one component of a test or ‘test method’. Good performance of a test system does not imply good functioning of a test method. Acceptability criteria for test systems (e.g. at least 75% of the differentiated cells staining positive for nestin under control conditions) are different from acceptability criteria for the test method using the test system (e.g. inhibition of differentiation by a specified positive control by at least 35%, and alteration of normal differentiation by a defined negative control by less than 10%) Test method A procedure based on a test system, used to obtain information on the biological effects of a substance. A toxicological test method consists of four major components (i.e. test system, exposure scheme, endpoint, prediction model), and it produces a test result (information regarding the ability of a substance or agent to produce a specified biological effect under specified conditions). The term is used interchangeably with “test” and “assay” in the literature. A test method can have several analytical endpoints. Prediction model

Author Manuscript

Prediction model: a formula or algorithm (e.g., formula, rule or set of rules) used to convert the results generated by a test method into a prediction of the (toxic) effect of interest. Also referred to as decision criteria. A prediction model contains four elements: (1) a definition of the specific purpose(s) for which the test method is to be used; (2) specifications of all possible results that may be obtained, (3) an algorithm that converts each study result into a prediction of the (toxic) effect of interest, and (4) specifications as to the accuracy of the prediction model (e.g., sensitivity, specificity, and false positive and false negative rates). In this context, ‘Data Interpretation Procedure (DIP)’ is of interest. It signifies any algorithm for interpreting data from one or more information sources. The output of a DIP is typically a prediction (e.g. prediction of skin sensitisation potential from peptide binding data and/or chemical structure).

Author Manuscript

Acceptance criteria Criteria defined before performing an assay to determine whether it is “valid”, i.e. whether the data can be used. Typical issues of acceptance criteria comprise: ‘has the actual run or plate of the test method functioned (e.g. are the endpoint values for PC and NC in the right range)’, ‘is the test method performing within the desired range of variability (e.g. are the standard deviations of PC and NC in the right range)’. Note: acceptance criteria can also be defined for an ‘analytical endpoint’ or for a ‘test system’.


Aschner et al.

Page 32

Endpoint

Author Manuscript

The biological or chemical process, response, or effect assessed in a test system by a specific analytical method/assay, e.g. “viability” as measured by LDH-release, expression of a marker as measured by PCR, or beating of cardiomyocytes evaluated by an imaging system. Note that each endpoint may be assessed by different ‘analytical methods’. For instance, ‘viability’ may be assessed by LDH-release, resazurin reduction, cell counting or measurement of ATP. “Differentiation” may be measured by PCR quantification of a differentiation marker or by morphometry (e.g. beating of cardiomyocytes evaluated by an imaging system. Analytical endpoint

Author Manuscript

An endpoint of a test system (e.g. proliferation, or differentiation, or viability) may be quantified by different analytical methods (measurement endpoins). It is important to distinguish such analytical endpoints (referring to the methods used) from (test systems) endpoints that refer to the biological concept evaluated. In vitro system This term has various meanings in the litereature, i.e. it is little defined. It is sometimes used to signify a cell/tissue culture system used as the basis for the development of a test method. In this sense it corresponsds to a test system (as above). (Note: In biochemistry, the term is often used for cell-free systems, as opposed to cellular (living) systems. Cell culture assays, i.e. in vitro assays in a toxicological sense, are often called “in vivo systems” in biochemistry). Assay

Author Manuscript

This term is used in a broader or narrower sense depending on the field, similar to ‘test method’. In a narrower sense, ‘assay’ can refer to an analytical procedure (e.g. protein determination, PCR). In a wider sense, ‘assay’ is used interchangeably with ‘test method’. A classical example is the Ames assay, which comprises a complex test system of growing and plating bacteria under different conditions together with an analytical procedure based on the counting of colonies. Reference compounds and statistics Positive/negative control (PC/NC)

Author Manuscript

A PC is a compound or condition that triggers a response, i.e. a change of the endpoint from baseline in a predicted direction and to a certain specified extent. A NC for a ‘test method’ is a compound or condition that should not trigger a response, i.e. it should not change the endpoint from baseline. The performance of PC and NC can be used to define ‘acceptance criteria’ of a test. Endpoint-Specific Controls Chemicals known to reliably and consistently alter the endpoint of a test system at a mechanistic level. These are also referred to as ‘endpoint-selective controls’ or ‘mechanistic tool compounds’. This would the first set of compounds used during test system setup to obtain information on the biological/toxicological behaviour of the test system and its dynamic range. ALTEX. Author manuscript; available in PMC 2017 April 01.

Aschner et al.

Page 33

Training Set Chemicals

Author Manuscript

This set should include chemicals known (preferably from in vitro systems) to reliably elicit a response, or no response, with respect to the endpoint of interest. The goal of using this set is proof-of-concept that the test method can rapidly and efficiently screen moderate numbers of chemicals with reasonable predictivity. A training set of chemicals can be used to optimize an assay (test method), to set acceptability criteria, and to build a prediction model. Testing Set Chemicals This set would be used to validate and possibly improve the prediction model. For DNT, this set should include chemicals known to affect (and also some that definitely do not affect) in vivo developmental neurotoxicity endpoints The goal of using ‘testing set chemicals’ is also to demonstrate the ability to test larger numbers of chemicals.

Author Manuscript

General cytotoxicity (GC) The term is used when a compound triggers cell death that is not specific for the cell type used in the assay but would occur in most cells at the same concentration and within a similar time frame. For many test methods it is important to measure specific adverse effects that occur at concentrations below those triggering cell death in the test system. Therefore, the verification of test conditions not triggering GC is important for many tests. Unspecific controls (UC)

Author Manuscript

Unspecific controls (UC): often refers to compounds displaying GC. For some test systems, it is sufficient to work with PC and NC. For other test systems, it is important to demonstrate a difference between compounds that act specifically, and compounds that lead to changes of the endpoint because they trigger GC. For instance, a test may be designed to determine the metabolic fingerprint of cell cycle blockers. Such a test would require the examination of UC and the comparison of their profile with PC compounds. Highest non-cytotoxic concentration (HNCC) Highest non-cytotoxic concentration (HNCC): the highest concentration of a compound that does not trigger GC. The HNCC is important, as it allows the detection of specific adverse effects with highest likelihood. It defines the highest concentration to be used in test systems examining particular toxic effects independent of GC. Testing at concentrations higher than the HNCC may lead to artifacts. Replicates within one experiment

Author Manuscript

These are also called “technical replicates” and can take two different forms: A: the repeated performance of an analysis on the same sample, e.g. duplicate PCR, Western blot or FACS determinations. B: the determination of an endpoint from more than one culture well, with all these wells being incubated in parallel/on the same day/in the same experiment.


Aschner et al.

Page 34

Independent experiments

Author Manuscript

These are also called “biological replicates” and should not be confused with technical replicates in different dishes. A biological replicate is a separate experiment, i.e. on another day, with independent cell batches, new test solutions, etc. A biological replicate can comprise several technical replicates. Robustness/Ruggedness

Author Manuscript

Is a measure of a methods’ capacity to remain unaffected by small variations in method parameters and environmental conditions. Testing of robustness provides an indication of a test’s reliability during normal usage. Sometimes a distinction is made between robustness and ruggedness. The latter focuses on the reproducibility of the test results obtained for identical samples under normal test conditions that underlie unintentional changes (room temperature, source of human sample material, lot variation of reagents, operator-dependent variables, weather conditions, etc.). Robustness testing would explore the insensitivity of a test to deliberate variations in the test environment or setup (incubation time, temperature, cell passage number, sample storage, cell density, type of culture dish, etc. Dynamic Range Determination of the extent of measurable change that can be detected for a DNT endpoint and whether both increases and decreases from untreated control can be measured. Test concepts Fundamental biological process

Author Manuscript

In the context of DNT, this refers to ‘fundamental neurodevelopmental process’. These processes include precursor cell proliferation, neuronal and glial cell differentiation and apoptosis, synaptogenesis and myelination, and are also termed ‘key biological processes’ or ‘key neurodevelopmental events’. They need to be distinguished from signalling events or more basic mechanisms, in that ‘fundamental biological processes’ represent a higher (superordinate) level of organization, that comprises many signaling mechanisms and targets of molecular intiating events. They are ‘fundamental’, as failure of any of them may result in DNT. Importantly, these processes can be modeled using in vitro test systems, and each such ‘test system’ has the advantage of capturing (identifying) many different toxicants acting by different molecular mechanisms. Note: fundamental biological processes are not to be confused with key events (KE) in an AOP. Molecular initating event (MIE) and key events (KE)

Author Manuscript

A molecular initiating event is the initial point of contact between a chemical and a specific biomolecule that results in a cascade of key events (KE) leading to an adverse outcome. Adverse outcome pathways (AOPs) Adverse outcome pathways (AOPs): conceptual constructs that link a molecular initiating event (MIE) to an adverse outcome at the level of the whole organism. The AOP links existing knowledge along one or more series of causally connected key events (KE) ALTEX. Author manuscript; available in PMC 2017 April 01.

Aschner et al.

Page 35

Author Manuscript

between two points — a molecular initiating event (MIE) and an adverse outcome (AO). AOP are not compound-specific, but a theoretical construct applicable to multiple compounds. Toxicity endophenotypes (TEP) Altered functional or structural connectivity or responsiveness of specific regions of the nervous system as a consequence of exposure to xenobiotic(s). TEP represent the level of organization that links in vitro test systems for fundamental biological processes to apical DNT endpoints in vivo (exophenotypes). Integrated Approach to Testing and Assessment (IATA)

Author Manuscript

Integrated Approach to Testing and Assessment (IATA): an approach based on multiple information sources used for hazard identification, hazard characterization and/or safety assessment of chemicals. An IATA integrates and weighs all relevant existing evidence and guides the targeted generation of new data, where required, to inform regulatory decision-making regarding potential hazard and/or risk.

Author Manuscript Author Manuscript ALTEX. Author manuscript; available in PMC 2017 April 01.

Aschner et al.

Page 36

Author Manuscript

Box 2 Criteria used to select chemicals as positive DNT controls Note: the letters refer to the superscripts in table 4.


a.

At least two peer-reviewed papers from two independent groups of investigators reporting evidence of DNT. For human data: only positive meta-analysis of multiple studies.

b.

DNT evidence from in vivo mammalian models with exposures during gestation and/or lactation (either direct pup exposure or exposure to the dam) prior to weaning; robust human epidemiological data also considered.

c.

Exposure is to the test chemical itself and not to formulations, its metabolites or mixtures (Shafer et al. 2005, Shafer and Crofton, 2011).

d.

Outcomes were neurobehavioral, neurophysiological, functional (including pharmacologic responses), brain anatomic or pathology findings not due to acute effects of exposure. Findings based solely on neurochemical, gene expression or biochemical endpoints were excluded from consideration.

e.

The statistical unit (e.g. individual pups or litters) is reported. For animal studies with gestational or early postnatal exposure (either lactational or direct dosing), the litter is the experimental unit (De Sesso et al., 2009, Holson et al., 2008). Violation of this criterion was accepted but flagged, as it was not always possible to distinguish because of poor study design and/or poor reporting standards.

f.

Minimal sample size is reported and is at least n=6 (i.e. 6 litters/dose group for gestational or early postnatal exposure studies).

g.

Studies were not included if the route of exposure was intracerebral injection. Preference was given to studies using human-relevant routes of exposure (i.e. oral, dermal, inhalation).

h.

Studies should be based on at least 2 dose-levels; some violation of this criterion was allowed but flagged, since single dose studies in one publication may often build on previous experience of the group with multiple doses,


Aschner et al.

Page 37

Author Manuscript

with subsequent studies based on the most appropriate dose.

Author Manuscript

i.

Relationship of maternal toxicity versus DNT: Ideally, DNT should occur at lower doses than maternal toxicity. Studies in which maternal toxicity occurred at the same dose as DNT, or where this was not reported, were flagged.

j.

Relationship between DNT and general toxicity: ideally, DNT should occur at lower/same concentrations than general toxicity. Studies in which general toxicity/ mortality occurred at the same concentration as DNT, or in which this was not reported, were flagged/ highlighted. Studies where this relationship was not reported were also flagged.


Aschner et al.

Page 38


Figure 1. Representation of the key events of neurodevelopment at the cellular level

Several fundamental neurodevelopmental processes are absolutely necessary for nervous system development, and therefore well-conserved across species. Moreover, the processes known from in vivo studies can be relatively faithfully modeled in vitro. It is assumend that DNT exert their toxicity, because they disturb at least one of these processes. Therefore, disturbances of the processes depicted here are KE of AOP relevant for DNT.


Aschner et al.

Page 39


Figure 2. Toxicity endophenotypes

Author Manuscript

For development of relevant model systems, we need approaches for linking the observable DNT effect (= exophenotype; see red box) triggered by a xenobiotic to effects that this compound has in in vitro test systems (yellow circles). Toxicity endophenotypes (orange box) form the conceptual link between what is observed in man or experimental animals and on what test systems model. They are a description of the altered biological state of the nervous system (e.g. neuronal disarray in the frontal cortex) in vivo that causes the externally observable DNT phenotype (e.g. reduced IQ). Thus, ‘toxicity endophenotypes (TEP)’ describe the altered functional or structural connectivity or responsiveness of parts of the nervous system, triggered by xenobiotics. The TEP results from the disturbance of one or several fundamental biological processes (e.g. neurite growth). Notably, there may be a delay or lag of years between disturbance of a process by a chemical and the observation of DNT effects (dashed arrows linking processes and TEP). Both the setup of model systems and the characterization of tool compounds to validate such systems requires that we establish the following connections: (1) exophenotype to TEP (the exophenotype is the only robust and relevant starting point for identification of DNT compounds known at present); (2) association of TEP with disturbed biological process(es) that led to the TEP; (3) link of in vitro test system endpoint to prediction of a disturbed biological process in vivo. The fundamental biological processes as such (but not the TEP) may be modeled by alternative test systems. Thus, the test systems are inspired by the biological processes (green arrows),


Aschner et al.

Page 40

Author Manuscript

but the outcome of test systems predicts to some extent certain TEP (e.g. inhibited neuronal migration predicts neuronal disarray and/or a deficit in neuronal number in some brain region). In this sense, TEP represent the level of organisation that links in vitro test systems for fundamental biological processes to apical DNT endpoints (exophenotypes).

Author Manuscript Author Manuscript Author Manuscript ALTEX. Author manuscript; available in PMC 2017 April 01.

Aschner et al.

Page 41

Tab. 1

Author Manuscript

Examples of events relevant for adverse outcome pathway (AOP) linking exposure to DNT chemicals to human toxicity An AOP represents a series of measurable key events (KE) with biologically plausible connections. They connect a molecular initiating event (MIE) to an adverse outcome (AO) in an individual. The AOP is a concept that provides a framework for organizing knowledge about the progression of toxicity events across scales of biological organization. Here examples are given for MIE, for KE (on the cellular and organ level), and for AO, i.e. the manifestation relevant for man, that may be triggered by DNT chemicals. The cellular KE correspond to fundamental neurodevelopmental processes as detailed in Fig. 2 Molecular initiating events (MIE) •

Author Manuscript

•

•

•

Author Manuscript

•

Key events (KE) – cellular responses

Key events (KE) – organ responses

Modulation of the function of ion channels;

•

Neural precursor proliferation;

•

migration;

inhibition of assembly or disassembly of cytoskeletal elements;

•

gliogenesis;

•

neuronal differentiation;

•

neurite growth (axons, dendrites);

inhibition of key enzymes (e.g. acetylcholine esterase or receptor tyrosine kinases); inhibition of the mitochondrial respiratory chain;

•

synaptogenesis;

•

oligodendrogenesis;

•

myelination;

•

programmed cell death;

•

neuroinflammation;

•

etc.

inhibition of transporters on the cell membrane or organellar membranes;

Author Manuscript

•

inhibition or stimulation of nuclear receptors;

•

inhibition of cell-cell or cell-matrix contacts;

•

inhibition of DNA synthesis;

•

modulation of epigenetic processes (e.g. histone modifications or DNA methylation);

•

etc.


•

S. n dop neu deg

•

Hip den neu

•

Hy in p wh

•

liss

•

mic

•

hol

•

alte pat

•

atte inh

•

alte ser reg

•

alte sei trea

•

etc

Aschner et al.

Page 42

Table 2

Apical in vivo endpoints of DNT translated to DNT endpoints in vitro

Author Manuscript

In vivo studies use various methods to evaluate DNT. These can be roughly classified as anatomical measures (e.g. morphology, histopathology) or as functional measures (e.g. motor, sensory and cognitive function). These methods assess various outcomes (e.g. malformations detected by anatomical measures) or changes (increase/decrease) in functional parameters. Each of these outcomes derives from changes in cellular biology (e.g. altered apoptosis, cell migration or cell proliferation may lead to size differences of brain regions). The cell biological changes may be modeled by in vitro or alternative test methods. Methods in vivo

Outcome

Gross morphology

Brain measures↑↓ Brain parts missing Malformation

Necrosis Pyknosis Neuronal degeneration Astrocytosis Layer thickness ↑↓

Histopathology

Author Manuscript

Layer thickness ↑↓ Morphology

Morphometry

Learning/Memory/Motor Activity

↑↓

Cell Biological Causes → → →

, , ,

→ → → → →

,

→ →

, ,

,

, GFAP content ,

→ → → →


, ,

,

Aschner et al.

Page 43

Table 3

Tool compounds/endpoint-specific controls for DNT test systems

Author Manuscript

Assays were classified according to the basic biological process they are modeling (left column). The literature was then screened for compounds that elicited robust positive responses in respective in vitro test systems. These compounds were classified according to their inhibiting or activating effect on the baseline or control readout. For compounds that interfere with cellular differentiation, this one-dimensional classification was not attempted. For practical purposes (choice of positive controls useful during assay setup), the table contains not only classical endpoint-specific controls but also chemicals/toxicants with unclear mode of action, but with a robust effect on the targeted endpoint. They were considered useful to evaluate the technical performance of the test system with respect to the endpoints measured. For each compound, the original literature documenting its effect on the targeted endpoint is indicated.


Inhibitory

Stimulatory

Migration

methylmercury7,8, PP27,8, AG14788, PD980598, SU66568, SP6001257, pertussis toxin7, lead acetate7, triadimenol7, thimerosal7, semaphorin3A7, valproic acid7, CK-6667, cytochalasin D7, 3methylcholanthrene9, 7NI10, ODQ10

albumax7, phorbol myristate acetate (PMA)8

Proliferation

aphidicolin11,12,13, cadmium11,12,13, cytosine arabinoside11,12,13, 5-fluoroacil11,12,13, methylmercury13

epidermal growth factor4

Synaptogenesis

mevastatin15, potassium chloride15

Network activity

bisindolylmaleimide16

domoic acid17

Neurite outgrowth

methylmercury14,18,19,20,21, U012614,18,19,20, bisindolylmaleimide I14,15,18, lithium14,15,20, sodium orthovanadate20,22,23, retinoic acid14,18, brefeldin A20, flavopiridol20, cycloheximide2, paraquat2, diquat2, rotenone2, nocodazole2, colchicine2, vincristine2, narciclassine2

Y-2763220, HA-10772, blebbistatin2

Oligodendrocyte differentiation

PBDE-9924, PBDE-4724

thyroxin25, PCB 11825

Differentiation (compounds known to alter this process (adversely) in one of many possible ways)

methylmercury1,2,3,4, mercury chloride5, valproic acid2,3, trichostatin A3, retinoic acid6, lead acetate6, cyclopamine6, bone morphogenetic protein (BMP)43

The numbers behind the compound refer to the literature references as follows:

1

(Zimmer et al., 2011b),

2

(Krug et al., 2013b),

3

(Balmer et al., 2012),

4

(Moors et al., 2009),

Author Manuscript

5


6

(Zimmer et al., 2011a),

7

(Zimmer et al., 2012),

8



Aschner et al.

Page 44

9

(Gassmann et al., 2010),

10 11

Author Manuscript

12 13

(Tegenge et al., 2011),

(Mundy et al., 2010), (Culbreth et al., 2012), (Breier et al., 2008),

14

(Harrill et al., 2011a),

15

(Harrill et al., 2011b),

16

(Robinette et al., 2011),

17 18 19

Author Manuscript

20 21 22 23 24 25

(Hogberg and Bal-Price, 2011),

(Radio et al., 2008),

(Radio et al., 2010),

(Stiegler et al., 2011), (Parran et al., 2001), (Harrill et al., 2010), (Mandell and Banker, 1998), (Schreiber et al., 2010), (Fritsche et al., 2005).


Aschner et al.

Page 45

Table 4

Suggestions for negative tool compounds


A set of potential negative controls has been assembled, and experience from multiple assays will be needed to further refine this list. Although absence of activity cannot be proven, compounds with a very high likelihood to not affect DNT assays are found amongst sugar derivatives, solvents and polymeric compounds that do not enter cells. These types of relatively trivial negative controls mainly provide an indication of assay robustness and background noise levels, but do not provide much information regarding assay specificity. Another group of potentially negative control compounds are those with defined pharmacologic effects or other measurable bioactivity that are unlikely to trigger DNT or to affect fundamental neurodevelopmental processes. However, compounds for which this information is known are not available for every test system. Notably, any compound has the potential to affect biological systems at high enough concentrations. Therefore, specific compounds are useful as negative controls only if used at appropriate concentrations. This may be the concentration known to be bioactive in other systems (e.g. clinically-observed plasma levels for drugs), the highest non-cytotoxic concentration or the highest concentration used for any positive control (e.g. 100 μM – 1 mM), as higher chemical concentrations are unlikely to occur in any in vivo situation. Note that compounds like nicotine may be good negative controls for some assays, e.g. cell migration, but endpoint-selective positive controls for other assays, e.g., neural network assays. Importantly, the absence of a drug’s specific target in a test system (e.g. warfarin), does not mean that there is not another, less characterized (or unspecific) target, that still leads to effects on test endpoints.


Compound

Comments

Literature

Anthracen

Polycyclic aromatic hydrocarbon; may act via Ah receptor, but has no target in many human DNT/NT test systems

1

3-Imino-propionitrile

Neurotoxicant, requiring metabolic activation. Low toxicity if test system lacks activating enzymes

2

Metoclopramid, amitryptilin, ibuprofen, metoprolol, sumatriptan, amoxicillin, diphenhydramine

Drugs that are acceptable during pregnancy

10

Pomalidomide

Thalidomide analog, no DNT up to 200 μM

3

Omeprazole/warfarin

Drugs with primary target only in stomach/liver; low likelihood to have DNT effects

4, 5

Captopril, dabigatran

Drugs with extracellular targets

–

Solvents: dimethylformamide, DMSO, glycerol

Generally low toxicity up to mM range

–

Sugar (derivatives): sorbitol, lactose, mannitol, glucosamine, diethylene glycol

No pronounced bioactivity, sometimes not entering cells, tolerated to mM level;

Belongs to “trivial” controls (low usefulness for specificity calculations) with solvents

–

Glyfosate

Pesticide tested negative for DNT; low cytotoxicity

–

Dinotefuran

Neonicotinoid pesticide without DNT effects in many systems (may however effect neuronal network assays)

6

Fipronil

Pesticide tested clearly negative for DNT; may be cytotoxic at > 10 μM; may have indirect effects through cramp induction (zebrafish)

7

Deprenyl

Antidepressant/parkinsonian drug, inhibitor of monoamine oxidase-B (1 mM range)

–

Acetaminophen/paracetamol

Negative in most systems up to mM levels, but has been discussed as in vivo DNT toxicant

8, 9

Saccharin

Artificial sweetener, very low toxicity

–


Aschner et al.

Page 46

Author Manuscript

Compound

Comments

Literature

Trolox, zVAD-fmk

Water-soluble vitamin E analog; caspase inhibitor (usable at 100 μM)

–

Deferoxamine mesylate

Iron chelator, tolerated at mM levels

–

Furosemide, verapamil, levetiracetam, statins, seroquel, naloxon, atropine. ursodeoxycholic acid, tiotropium

Drugs with low likelihood to affect DNT test systems, due to their well characterized side effects and mode of action (may have direct effects on neural networks, though)

RU38486, propylthiourcil, testosterone…

Hormone modifiers little relevant to in vitro DNT test system targets

The numbers behind the compound refer to the literature references as follows:

1

(Pei et al., 2015),

2

(Ryan et al., 2016),

3

(Mahony et al., 2013),

4

(Gill et al., 2009),

Author Manuscript

5

(Ekman et al., 1985),

6

(Sheets et al., 2016),

7

(Krug et al., 2013a),

8

(Burdan, 2003),

9

(Reel et al., 1992),

10

(Niebyl and Simpson, 2008).


–

Aschner et al.

Page 47

Table 5

Compounds triggering DNT in vivo


An initial list of compounds was collected from the literature by way of subject expert suggestions. This list was intended to be exemplary and not exhaustive or even complete. In a second step, each compound was scrutinized for published literature supporting its DNT activity. The criteria described in Box 2 were applied to evaluate supporting literature (supplementary excel file). As an additional criterion, we used ‘strong evidence for DNT effects in humans’ as documented by well-recognized meta-analysis or well powered studies (column ‘Hu’, for human evidence). Compounds were retained in the list when at least two publications from two different laboratories in support of their DNT activity were identified. Published studies were categorized into one of four certainty groups: a) animal study that meets all criteria as described in Box 2 (score 3); b) study describes human data with statistically representative populations or study represents meta-analysis of human findings (score: 3), c) animal study in which one criterion is not met (score: 2); d) animal study in which 2–4 criteria were not met (score 1). For the classification of papers, criteria 5 and 8 as described in Box 2 were not included, but they are indicated for transparency. For the assessment of the certainty of the developmental neurotoxic effects of the selected compound, the scores were averaged. Compounds with a score of 2.5 or higher are presented in green, compounds with a score of 1.5–2.5 are presented in light green. Compounds with lower scores were eliminated. The superscript numbers (explained in Box 2) for each publication indicate the selection criteria not met. The comment field gives an indication on the endpoints used in the studies. If different types of endpoints were used they are indicated in the sequence of the listed publications, separated by semicolon.


Compound

Reference

Additional comments

Hu

Arsenic

5; 6e,f; 7f,h

Behavior

2

Cadmium

8e,i; 9

Behavior

Chlorpromazine

10e; 11f,h,i

Behavior; seizure threshold

Chlorpyrifos

12; 13

Brain cholinesterase inhibition; brain weight and morphometry

Cocaine

14; 15h,j; 16h,j

Human; behavior + morphology

Dexamethasone

17e,f,i; 18e,f

Behavior; behavior, brain chemistry; human: cortisol values, stress response

Diphenylhydantoin (Phenytoin)

20i;21i

Behavior; behavior, eye opening

Domoic acid

22e; 23; 24e,f,h,i

Conditioned place preference, activity; memory, behavior; neurochemistry

Ethanol

25; 26i,j; 27

Human: behavior; behavior, learning; attention; human: morphology

Haloperidol

29e,f,h; 30h; 31h

Behavior/cognitive

Heroin

32e,h,i; 33e,h,i; 34h,i

Human: behavior

4

Hexachlorophene

36e,h,j; 37

Human: neuropathology; vacuolation of brain white matter

35

Ketamine

38e,j; 39; 40h

Motor activity, learning, memory; increased apoptosis; behavior, spatial learning

Lead

41f,i; 42e,i; 43f,I,j

Human; behavior; mRNA expression, brain enzymatic activity; brain chemistry

Lindane

44f,I,j; 45e,h

Behavior

MAM

49h,j, 50f, 51h,i,j

regional brain weight; increased innervation, neurochemistry; brain morphometry

Maneb

52e,i, 53h

Behavior; behavior, morphology (in vivo cell count)


3

19

4, 28

1

Aschner et al.

Page 48


Compound

Reference

Additional comments

Hu

Manganese

46e,f,h,i, 47e,h, 48f,j

Behavior, brain chemistry

3

MDMA

54, 55h,i

Behavior; neuropathology; human: cognition; human: mental/motor development

56, 57

Methanol

58h,I,j, 59h, 60e,h,i

Behavior

Methyl mercury

61, 62e,f,h, 63i

Human; behavior; behavior; neurobiochemistry, transcriptomics

MPTP

64e,h, 65e

Behavior, brain neurochemistry; behavior

Nicotine

66e,h, 67h

Behavior

Paraquat

68e,f, 69h

Behavior; brain neurochemistry

PBDE

70e, 71h, 72e,h

Behavior; behavior, pharmacologic challenge; electrophysiology

3

PCB

73, 74

Human: Behavior, brain morphometry; behavior

1

Perfluorate – PFOA

75e,f, 76e,h

Behavior

Perfluorate – PFOS

77e,i,j, 78e,i, 79e,j

Hippocampus structure; behavior, motor activity, learning, memory,

Terbutaline

83h, 84e

Behavior; behavior, neuroinflammation

Toluene

85e, 86e

Behavior; brain weight

Trans retinoic acid

80i, 81i,j, 82h

Behavior; behavior; motor coordination, learning, brain morphology

Triethyl-tin

87j, 88j

Behavior, brain cell count; brain weight, myelin basic protein

VPA valproic acid

89, 90e

Behavior

References:

1

(Grandjean and Landrigan, 2006),

2

(Tolins et al., 2014),

3

(Grandjean and Landrigan, 2014),

4

(Yolton et al., 2014),

Author Manuscript

5

(Martinez-Finley et al., 2009),

6

(Chattopadhyay et al., 2002),

7

(Rodriguez et al., 2002),

8

(Baranski, 1984),

9

(Ali et al., 1986),

10 (Robertson et al., 1980), 11 12 13

Author Manuscript

14 15 16 17

(Golub and Kornetsky, 1975),

(Johnson et al., 2009), (Maurissen et al., 2000), (Mactutus et al., 2011),

(Kabir et al., 2014), (Lu et al., 2012), (Hossain et al., 2008),


1

1

Aschner et al.

Page 49

18

(Benesova et al., 1999),

19 20

(O’Connor et al., 2013),

Author Manuscript

(Weisenburger et al., 1990),

21

(McCartney et al., 1999),

22

(Doucette et al., 2003),

23

(Levin et al., 2005),

24

(Dakshinamurti et al., 1993; Lucchi et al., 1983),

25

(Oshiro et al., 2014),

26

(Lucchi et al., 1983),

27

(Brys et al., 2014),

28 (Fryer et al., 2012),

Author Manuscript

29 30 31 32 33 34 35

Author Manuscript

40 41

(Rosengarten and Quartermain, 2002),

(Yanai et al., 1992), (Wang and Han, 2009), (Shuman et al., 1974),

37

39

(Wolansky et al., 2004),

(Lasky et al., 1977),

36

38

(Watanabe et al., 1985),

(Ulsamer et al., 1975), (Itahashi et al., 2015),

(Fredriksson et al., 2007), (Paule et al., 2011), (Zhao et al., 2014), (Petit et al., 1992),

42

(Hu et al., 2008),

43 (Reddy et al., 2007), 44

(Johri et al., 2007),

45 46

Author Manuscript

47 48

(Rivera et al., 1990), (Lown et al., 1984),

(Kristensson et al., 1986),

(Deskin et al., 1981),

49

(Sullivan-Jones et al., 1994),

50 (Cattabeni et al., 1989),


Aschner et al.

Page 50

51

(de Groot et al., 2005),

52

(Sobotka et al., 1972),

53

Author Manuscript

(Thiruchelvam et al., 2002),

54 55 56

(Broening et al., 2001),

(Thompson et al., 2012), (McElhatton et al., 1999),

57

(Singer et al., 2012),

58

(Stern et al., 1997),

59

(Infurna and Weiss, 1986),

60

(Aziz et al., 2002),

61

(Elsner et al., 1988),

Author Manuscript

62

(Sakamoto et al., 2002),

63

(Radonjic et al., 2013),

64 (Ochi et al., 1991), 65 (Fredriksson et al., 1993), 66 67 68 69 70

Author Manuscript

71 72 73 74

(Levin et al., 1993),

(LeSage et al., 2006), (Fredriksson et al., 1993), (Thiruchelvam et al., 2002),

(Viberg et al., 2003), (Dufault et al., 2005), (Dingemans et al., 2007),

(Yang et al., 2009), (Sable et al., 2006),

75

(Johansson et al., 2008),

76

(Onishchenko et al., 2011),

77

(Zeng et al., 2011),

78

(Butenhoff et al., 2009),

79

(Johansson et al., 2008),

Author Manuscript

80

(Nolen, 1986),

81 (Holson et al., 1997), 82 83

(Coluccia et al., 2008),

(Owens et al., 2011),


Aschner et al.

Page 51

84

(Zerrate et al., 2007),

85 86

Author Manuscript

87

(Hass et al., 1999),

(Burry et al., 2003), (Freeman et al., 1994),

88

(O’Callaghan et al., 1983),

89 90

(Vorhees, 1987), (Schneider and Przewlocki, 2005).

Author Manuscript Author Manuscript Author Manuscript ALTEX. Author manuscript; available in PMC 2017 April 01.

Alternative methods of attenuating Salmonella species for potential vaccine use.

International STakeholder NETwork (ISTNET): creating a developmental neurotoxicity (DNT) testing road map for regulatory purposes.

Development of a list of reference chemicals for evaluating alternative methods to in vivo fish bioaccumulation tests.

Validated methods for degrading hazardous chemicals: some halogenated compounds.

Alternative and efficient extraction methods for marine-derived compounds.

Method for detecting the reactivity of chemicals towards peptides as an alternative test method for assessing skin sensitization potential.

Trithiols and their arsenic compounds for potential use in diagnostic and therapeutic radiopharmaceuticals.

Organotin compounds and their use for insect and mite control.

Criteria for judging the relative toxicity of chemicals from developmental toxicity data: a workshop summary.

Selection and use of mass reference standards.

Aroma chemicals: Reference sources for perfume and flavour ingredients with special reference to cinnamic aldehyde.

Untapped Potential: Animal care and use programs adopt alternative training and recruitment strategies for individuals with intellectual and developmental disabilities.

Aquatic toxicity testing for multicomponent compounds with special reference to preparation of test solution.

Indications and selection criteria for cardiac transplantation.

Zebrafish as an Alternative Vertebrate Model for Investigating Developmental Toxicity-The Triadimefon Example.

Commentary: peer review and incomplete reference lists.

Comparative toxicity of six test chemicals to lettuce using two root elongation test methods.

Alternative agents in refractory chronic urticaria: evidence and considerations on their selection and use.

Screening methods for assessment of biodegradability of chemicals in seawater--results from a ring test.

Criteria of radon risk of territories and methods for their determination.

Classification of chemicals as sensitisers based on new test methods.

Bayesian methods for prevalence estimates from incomplete administrative lists.

The Warsaw Proposal for the Use of Extended Selection Criteria in Liver Transplantation for Hepatocellular Cancer.

Criteria for the selection of restoration materials.